Accidental falls among the elderly present serious health risks and are a significant concern, particularly for individuals living alone. Annually, approximately 2.8 million seniors require emergency medical attention due to fall-related injuries, highlighting the urgent need for effective fall detection and response mechanisms. While video-based fall detection systems tend to be more expensive than wearable solutions, their ability to integrate with smart home technologies enhances their practicality and real-time monitoring capabilities. This review systematically examines video-based fall detection methodologies, assessing their effectiveness, challenges, and constraints across different processing stages. Furthermore, we provide a comparative analysis of state-of-the-art techniques, identifying key advancements and potential areas for future research to improve reliability and accuracy.

The health sector has witnessed phenomenal advances over the last several years due to numerous technological innovations that have significantly impacted almost all medical processes. Among these innovations is robotics, which has, over the recent years, proved a keystone in revolutionizing how healthcare amenities are structured. This chapter therefore extensively discusses the many roles that robots occupy in providing healthcare, ranging from telemedicine to supporting surgical procedures. Part of the discussion includes the analysis of varying types of robots, how they are currently being utilized in the real world, and their significance on a patient’s well-being. The chapter also underscores the creative models of India, including SSi Mantra, the first surgical robot project that has revolutionized the field of medical technologies. Particularly in the wake of the COVID-19 pandemic, disinfecting robots have significantly aided frontline workers in keeping the flow of healthcare delivery. Even with the unveiling of robotics in the health sector, there are challenges. This chapter, therefore, portrays the reciprocity between robotics and healthcare, indicating the evolving nature of the medical fields and the bright future robotics holds for the sustainability of healthcare delivery

Accidental falls and their consequences are critical concerns for elderly people. Fatal injuries, when delayed in treatment, can lead to severe outcomes. Fall detection systems are crucial for the timely treatment of such injuries. Although sensor-based fall detection approaches are effective, video-based approaches are more useful because they assist in analyzing the fall scene and identifying the cause of the fall. However, privacy preservation is a major concern in video-based fall detection. The proposed system introduces a privacy-preserving mechanism that masks the identified human with a silhouette. A custom dataset, including 80 activities of daily living and 70 fall activities, is introduced. An LSTM variational autoencoder architecture is designed with a gradient clipping mechanism and a smooth variant of Adaptive Moment Estimation with Stochastic Gradient Descent (AMSGrad) optimizer to enhance the accuracy of fall detection. The reconstruction error between normal and fall activities is clearly identified with the help of a dynamic threshold. This results in a system performance that achieves accuracy, precision, and sensitivity of 99%, 97%, and 99%, respectively

Blindness is a person’s impairment to sight. Our sense of sight is responsible for most of the information from the five senses. We develop a prototype of our idea of a new person based on the initial meeting, taking cues from the tiny behaviors and continuously refreshing our understanding of that person. A blind person cannot perform these. For this reason, there is a need to design such a system that is capable of collecting fundamental information from the instant a new person appears and is portable for convenience of operation. The information we can extract includes but is not limited to-gender, age, and emotions at the moment of conversation. Hence, we have utilized the Haar cascade approach in conjunction with a face recognition module to detect faces using a Convolutional Neural Network (CNN) to train models that can do age detection, gender detection, and sentiment analysis. To make it portable, we have modified our models to run on edge devices such as Raspberry Pi, NVIDIA Jetson Nano, Intel Neural Compute Stick, Nano Pi, Etc. At its core, our project is about aiding people with visual impairments. We want to make sure they can gain a deeper grasp of the people they interact with, such as knowing their age, gender, and even how they are feeling. We are using cutting-edge technology and making it run steadily. so that it is easy for them to utilize

The Keyword-Based Video Retrieval System (KB-VRS) is a potential solution to organize and access vast amounts of video content. We have focused our research efforts on building a reliable and efficient KBVRS and this paper presents an in-depth review of our findings. Our solution leverages advanced technologies like Natural Language Processing (NLP) and Optical Character Recognition (OCR) to automate the analysis and indexing of video content based on user-defined keywords. The KBVRS extracts keywords and key phrases from video frames and audio transcripts, enabling efficient searching and retrieval of relevant video information, thus enhancing multimedia content management. Our goal is to overcome the limitations of manual tagging and classification by providing scalable and customized KBVRS for multimedia content organization and retrieval. The system caters to users, including educators, researchers, media professionals, and content creators. The user-friendly interface and intuitive search feature facilitate easy access and utilization of multimedia information. We have shown through rigorous experiments that our system is resilient and effective in retrieving relevant video content based on user queries. Our paper contributes to keyword-based video retrieval systems by laying the groundwork for future research in this rapidly evolving field. These insights pave the way for further exploration in this dynamic field. Our study empowers better decision-making processes in video content management

Integrating artificial intelligence (AI) is crucial for addressing challenges in healthcare, particularly in medical data analysis and drug recommendations. This paper presents two methodologies to improve healthcare decision-making using data mining and AI-driven drug recommendation systems. The first method employs data mining, sentiment analysis with the Vader tool, and Natural Language Processing (NLP) on extensive medical datasets from Hospital Information Systems (HIS). It accurately predicts diseases and offers personalized drug recommendations based on data insights, enhancing precision with a weighted average approach. The second method highlights AI’s importance in drug recommendation systems, addressing the challenge of staying updated on the latest treatments. We developed a system using NLP and Machine Learning (ML) algorithms to predict medical conditions and recommend drugs based on reviews and their usefulness. Simple symptom input provides individuals with information on their disease and helpful drugs. These methodologies significantly advance healthcare decision-making, with sentiment analysis capturing patient experiences. Experimental results demonstrate these methods effectively enhance healthcare decision-making, improving patient outcomes and efficiency

Digital images that are frequently encountered in day-to-day life can be easily tampered to mislead the information. One of the popular methods used to accomplish this unauthorized alteration is copy-move forgery. This paper presents a passive authentication scheme for copy-move forgery which uses discrete cosine transform (DCT) with block-wise peak-pixel-based block clustering. This scheme initially aims to obtain the features by implementing DCT on small, fixed image blocks and minimizes the size of feature vectors. The block-wise peak-pixel-based block clustering algorithm is used instead of the general lexicographic order technologies to enhance the detection precision. By comparing the feature vectors in each bucket, similar blocks will be obtained. Based on the experimental outcomes, the proposed scheme can detect multiple irregular and significant tampered regions. The duplicated regions detected in the distorted digital images can also be displayed by adding white Gaussian noise, Gaussian blurring and their mixed operations. By applying the above approach to the CoMoFoD - Image Database an accuracy of 99.99% has been achieved.

A deepfake is a computer-generated image or video that appears to be real but is a fabricated representation created to make an individual appear to be saying or doing something that did not occur. Deepfakes generate misleading or deceptive information by manipulating and superimposing faces onto pre-existing footage using artificial intelligence. This paper introduces a novel approach for deepfake detection through a combination of EfficientNet and Recurrent Neural Networks (RNNs). This method enhances detection efficiency by leveraging the hierarchical features acquired by EfficientNet and employing RNNs, specifically Long Short-Term Memory (LSTM) networks, to capture temporal dependencies. Application of this approach to the Celeb-DF dataset resulted in an accuracy of 99.98%.

The study focuses on the utilization of remote sensing data to analyze and detect deforestation patterns, with an emphasis on the extraction of key parameters such as vegetation cover change, forest loss, and land use dynamics. Various image processing methods, encompassing supervised and unsupervised classification, object-based image analysis, and change detection algorithms, are discussed in the context of their applicability to deforestation monitoring. The benefits and limitations of each technique are identified, highlighting the significance of choosing the most appropriate method based on the specific needs of the study area and the required level of accuracy. The paper also explores the incorporation of remote sensing data with geographic information systems (GIS) and other ancillary data sources to enhance the analysis and interpretation of deforestation patterns. The findings from this work contribute to the advancement of remote sensing image processing methods for deforestation monitoring, offering valuable insights for researchers and practitioners in the field.

Maintaining road infrastructure is essential to effective transportation systems and public safety. This research provides a new method for pothole depth estimation and automatic road crack detection using computer vision techniques. Our method utilizes convolutional neural networks (CNNs) for classifying road images into “With Crack” or “Without Crack” categories with high accuracy. Additionally, we employ image processing algorithms to detect and highlight cracks, providing insights into their lengths and percentages. Furthermore, we introduce a monocular depth estimation model to assess pothole depths, aiding in prioritizing road repair efforts. Experimental results demonstrate the effectiveness of our approach in accurately identifying road defects and estimating their severity. This research contributes to the advancement of intelligent infrastructure management systems, enabling proactive maintenance and ensuring safer roads for communities.

Waste collection is a necessary everyday task that must be carried out frequently in our society. An inefficient waste management system results in unsanitary conditions. With the rapid development of technology, the concept of a smart city has entered our lives, and the cleaning processes in the cities are supported by modern technologies. As cities move towards becoming smart, it is possible to improve waste collection systems. A key innovation in our system is the development of smart dustbins using Arduino MKR WiFi 1010 and ultrasonic sensors. These smart dustbins continuously monitor their fill levels and communicate this data in real time. This data is then processed by machine learning algorithms, optimizing collection routes based on factors such as distance, capacity, time taken, fuel consumed, and historical data. In the era of smart cities, the integration of IoT, machine learning, and smart dustbins presents an opportunity to revolutionize waste management. By minimizing collection time, reducing costs, and minimizing the environmental footprint, our system contributes to cleaner and more sustainable urban environments. This paper not only presents the conceptual framework but also provides insights into the practical implementation of smart dustbins, demonstrating the potential for transformative change in waste collection systems

This paper explores how machine learning is transforming the financial markets, specifically in relation to trading strategies and profit-maximizing investment choices. We examine machine learning’s influence on algorithmic trading, revealing how it shapes trading algorithms and guides investment strategies to maximize returns, commonly called "alpha." From there, we explore a wide range of financial applications of machine learning, such as fraud detection, portfolio optimization, and risk management. While highlighting the positive impacts of machine learning, we also acknowledge and address the challenges it presents in finance. Overall, this article shows how machine learning (ML) is changing the financial landscape and emphasizes the need for ethical use to maximize its advantages while minimizing its risks

This paper proposes a novel approach to image stitching and sharpening that makes use of powerful keypoint extraction and matching techniques. For precise keypoint detection, the Scale-Invariant Feature Transform (SIFT) technique is used, while the Random Sample Consensus (RANSAC) approach generates homography, which improves alignment across input pictures. Keypoint matching creates correspondences in overlapping regions, allowing for more accurate alignment. To increase image quality, the approach utilises a novel sharpening process that draws on localised information from aligned keypoints. The experimental results show that the approach is effective in producing more precise images and accomplishing accurate image stitching. This method is a major contribution to the area, with applications in various areas such as Mars rover research, computer vision, remote sensing, and general image stitching.

A drastic change in communication is happening with digitization. Technological advancements will escalate its pace further. The human health care systems have improved with technology, remodeling the traditional way of treatments. There has been a peak increase in the rate of telehealth and e-health care services during the coronavirus disease 2019 (COVID-19) pandemic. These implications make reversible data hiding (RDH) a hot topic in research, especially for medical image transmission. Recovering the transmitted medical image (MI) at the receiver side is challenging, as an incorrect MI can lead to the wrong diagnosis. Hence, in this paper, we propose a MSB prediction error-based RDH scheme in an encrypted image with high embedding capacity, which recovers the original image with a peak signal-to-noise ratio (PSNR) of ? dB and structural similarity index (SSIM) value of 1. We scan the MI from the first pixel on the top left corner using the snake scan approach in dual modes: i) performing a rightward direction scan and ii) performing a downward direction scan to identify the best optimal embedding rate for an image. Banking upon the prediction error strategy, multiple MSBs are utilized for embedding the encrypted PHR data. The experimental studies on test images project a high embedding rate with more than 3 bpp for 16-bit high-quality DICOM images and more than 1 bpp for most natural images. The outcomes are much more promising compared to other similar state-of-the-art RDH methods.

In today's world as there is a rapid growth of digitalization, searching and retrieving of relevant images from large datasets efficiently and accurately is challenging. There are two Image Retrieval Methods. These are Text-Based Image Retrieval and Content-Based Image Retrieval. In this manuscript, we are going to implement both of these methods together to retrieve similar restaurant images of the same name and belonging to either the same franchise or having the same features as the query image. CBIR can be used to identify images of similar color and features from dataset to given query image and OCR is used to identify the textual information present on images. For more accurate results we are adding a phonetic-based Fuzzy string matching algorithm to improve the efficiency of OCR as the names of restaurants may not always be in a formal style.

Hyperspectral imaging, with its high spectral resolution, provides valuable information for land cover classification and remote sensing applications. Leveraging this potential, we present a novel approach to land cover classification in hyperspectral imaging (HSI) through spectral-spatial fusion using deep learning techniques. This research integrates Singular Value Decomposition (SVD) as a preprocessing step to enhance spectral information and employs a custom 3D Convolutional Neural Network (CNN) model for spatial feature extraction. The proposed work effectively combines spectral and spatial characteristics, addressing the unique challenges posed by hyperspectral data. SVD, as a dimensionality reduction technique, optimizes spectral information for efficient processing. The 3D CNN model captures spatial patterns and dependencies, enabling improved land cover classification accuracy. We indicated the effectiveness of the spectral-spatial fusion technique on a diverse hyperspectral dataset, achieving considerable improvements in land cover classification accuracy compared to traditional methods. The fusion technique not only enhances the classification performance but also provides understandable features for remote sensing applications.

This book reviews research works in recent trends in blockchain, AI, and Digital Twin based IoT data analytics approaches for providing the privacy and security solutions for Fog-enabled IoT networks. Due to the large number of deployments of IoT devices, an IoT is the main source of data and a very high volume of sensing data is generated by IoT systems such as smart cities and smart grid applications. To provide a fast and efficient data analytics solution for Fog-enabled IoT systems is a fundamental research issue. For the deployment of the Fog-enabled-IoT system in different applications such as healthcare systems, smart cities and smart grid systems, security, and privacy of big IoT data and IoT networks are key issues. The current centralized IoT architecture is heavily restricted with various challenges such as single points of failure, data privacy, security, robustness, etc. This book emphasizes and facilitates a greater understanding of various security and privacy approaches using the advances in Digital Twin and Blockchain for data analysis using machine/deep learning, federated learning, edge computing and the countermeasures to overcome these vulnerabilities.

With recent advances in assistive technologies, improving spatial perception for the visually impaired has proven to be a major challenge. While several methods have been proposed, a holistic real-time solution remains the subject of intense research. In this paper, we present an innovative approach that combines computer vision, speech recognition, and artificial intelligence. At the core of our system is the YOLOv4 model integrated with OpenCV, which has been carefully optimized for robust object recognition in heterogeneous environments. A novel distance estimation algorithm is introduced that uses the focal length of the device's camera to infer the actual distance based on the dimensions of the detected object within the image. To enhance user interaction, we integrated Palm API from Bard AI, which provides rich auditory descriptions of the environment. This innovative paradigm incorporates other advanced function-alities, encompassing hazard and speed detection with dual OS mode. Furthermore, the model flexibly adjusts to diverse network bandwidths through optimized configurations. Our paper offers a new paradigm in the field of assistive technology and sets a benchmark for future efforts aimed at reducing the barriers created by visual impairments.

-

Pneumonia ranks among the world's major causes of mortality and is the greatest cause of death for young children. It is an infectious condition that can be fatal, affects one or both lungs and is brought on by harmful bacteria. An accurate and timely diagnosis is essential for managing and treating patients effectively. Radiotherapists with specialized training are needed to assess chest X-rays to diagnose pneumonia. Therefore, creating an automated approach to identify pneumonia would be advantageous to treat the illness, especially in isolated locations quickly. This project offers a novel method for improving chest X-ray image quality, which is then used in conjunction with machine learning approaches to increase the detection accuracy of pneumonia. Subtle details in X-rays can be seen much better using picture-enhancing techniques including sharpening, contrast stretching, and histogram equalization. A VGG net and a convolutional neural network (CNN) model that can accurately diagnose pneumonia is trained using this augmented image dataset. By bridging the gap between conventional X-ray imaging and sophisticated machine learning, the initiative offers a viable approach to the early and accurate detection of pneumonia. Early disease identification is greatly aided by medical imaging, and chest X-rays are a frequent method of identifying lung disorders like pneumonia. This project offers a novel method for improving chest X-ray image quality, which is then used in conjunction with machine learning approaches to increase the detection accuracy of pneumonia. Subtle details in X-rays can be seen much better using picture-enhancing techniques including sharpening, contrast stretching, and histogram equalization. A Convolutional Neural Network (CNN) model that can accurately diagnose pneumonia is trained using this augmented image dataset. By bridging the gap between conventional X-ray imaging and sophisticated machine learning, the initiative offers a viable approach to the early and accurate detection of pneumonia.

Life expectancy has improved with new-age technologies and advancements in the healthcare sector. Though artificial intelligence (AI) and the Internet of Things (IoT) are revolutionising smart healthcare systems, security of the healthcare data is always a concern. Reversible data hiding (RDH) is widely explored in the healthcare domain for secure data transmission and in areas like cloud computing, satellite image transmission, etc. Medical image transmission plays an important role in the smart health sector. In the case of medical images, a minute error in the reconstructed medical image can mislead the doctor, posing a threat to the patient’s health. Many RDH schemes have been proposed, but very few address from the view of medical images, and that too on high-quality DICOM images. The proposed AdaBPE RDH scheme is a solution for secure transmission of the patient’s health report (PHR) and other sensitive information with medical specialists. The scheme put forward a technique that maintains a good trade-off between the smooth pixels for maximum embedding in a block and a lossless recovery. Here, the cover medium employed to hide the patient’s sensitive information is an encrypted 16-bit DICOM image. The scheme processes the cover image as disjoint blocks of equal size, embedding the information adaptively within the encrypted blocks pertaining to the nature of the actual pixel values in the block through MSB prediction error methodology. The outcomes are evaluated on both the 16-bit DICOM images and 8-bit natural images, and the scheme is well poised with RDH goal of BER = 0, PSNR = ?, and SSIM = 1, achieving an average embedding of 5.7067 bpp on high-quality medical images and 1.6769 bpp on natural images. The experimental results prove advantageous and are better than other similar state-of-the-art schemes. © 2024 The Author(s). CAAI Transactions on Intelligence Technology published by John Wiley & Sons Ltd on behalf of The Institution of Engineering and Technology and Chongqing University of Technology.

Reversible data concealment is a strategy that quietly alters digital material to keep secret data while allowing the original digital media to be totally restored without any error after extracting the concealed information. Reversible data hiding is often used for the content identification of visual data such as photos, films, digital records, and so forth because of its expanding demand in fields like enforcement agencies, diagnostic imaging, scientific research, and so forth. Data hiding is a mechanism of securely conveying data within a multimedia container. Reversible data concealment is a way of hiding data by lightly changing digital media while permitting the original digital information to be completely restored without error if the hidden data is found. Data hiding is a secure method of sending data within a multimedia container. In uses such as medical image transmission, military, and forensics, it is unlawful to destroy the original cover. Data-concealing solutions are popular in these domains. After recovering the concealed data, reversible data hiding (RDH) may precisely rebuild the cover image. Without knowing the encryption key or the original content of the picture, further information can be inserted in the encrypted image. Reversible data-hiding techniques completely restore the original carrier when the secret encrypted data is removed. The classification of reversible data concealing schemes is determined by the manner of implementation. We will cover all of the ways based on Histogram Shifting, Difference Expansion, Prediction error-based approach, and Compression embedding in this paper’s review.

Due the industrial development many countries facing a shortage of farming land. Finding enough manpower to carry out the necessary agricultural activities is also a major concern. With the limited farming land and workforce, the agriculture sector is expected to produce more to meet the requirements of a large population. Technology can support the agriculture industry in various ways to improve productivity. The use of advanced technologies such as the Internet of Things, machine learning, robotics, blockchain, etc. in the agriculture industry to improve productivity more efficiently is known as smart agriculture. This paper gives an overview of various methods in practice for smart agriculture and the current challenges in this area to which researchers can contribute. A detailed analysis of the existing methods is done in this study. © 2023 IEEE.

A chatbot is one kind of software developed to enable interaction using natural language processing between a user/human and a system. Many systems today are equipped with chatbots for interpreting user questions and providing the right answers in a fast and efficient manner. In this paper, we propose a method to develop a chatbot for academic-specific data access for the college, through a simple mobile application. The proposed work aims to build an Artificial Intelligence (AI) chatbot to access information related to students' inquiries towards their academic scores, attendance status, and other general and college-specific data. This chatbot application eliminates any human intervention in the process of student data retrieval. Furthermore, the application guarantee time optimization by not relying on having an active internet connection and reduces delay in navigation through the existing conventional website. Questions can be asked by students to the chatbot at any time of the day and quick responses are generated. Students can have conversations simultaneously with the help of a chatbot which works 24×7. Through natural language understanding, the chatbot extracts meaning, intents, entities, and context from conversational textual inputs.

Automated content moderation systems have become essential for maintaining a safe and healthy online environment with the ever-increasing amount of video content on the internet and social media platforms. Individuals are likely to see videos containing flashing lights that can trigger seizures, nudity, or videos that may evoke extreme emotions like anxiety, frustration, or despair. In this paper, we propose a comprehensive content moderation solution that incorporates several features. The system features epileptic seizure recognition, emotional dysregulation prevention, and kid-safe mode. Epilepsy is a chronic neurological disorder affecting millions worldwide. Various factors can trigger seizures, including flashing lights, which are prevalent in video content. The system analyzes videos for trigger segments and skips them to non-trigger segments, mitigating the risk of seizures in individuals with epilepsy. The proposed algorithm uses a sophisticated analysis of the differences in luminosity between each video frame. It identifies areas that represent potential seizure-inducing segments due to their high density of big changes in luminosity. Similarly, when a video triggers strong emotions, the emotional dysregulation prevention mechanism recognizes it and alerts the user with an emotional summary of the video. The emotional dysregulation algorithm uses a deep learning model for facial emotion identification based on the VGG16 architecture. In kid-safe mode, the nudity detection algorithm uses CNN architecture, recognizes explicit content, and blocks users, especially children, from seeing it. The system displays notifications and alerts to parents to establish restrictions on the content their children can access. The proposed mechanisms prevent potential harm to users with specific vulnerabilities and provide parents with tools to ensure a safe online streaming environment for their kids.

In the current digital age, we see that reviews from social networks or microblogging significantly influence one's decision-making. It has been observed that a large number of reviews and comments regarding educational institutions are placed on these sites. As we all know, many new platforms have been formed over time, increasing the activity of users to voice their opinions on these platforms. As a result, users, particularly students, are more likely to become perplexed while trying to conclude. To address this issue, this paper looks at the analysis of a few platforms as a source of getting the reviews/comments of Universities, and with the help of sentiment analysis, we get an overview of a specific universities' position in the market, which includes certain parameters such as quality of faculty, mess, infrastructure and so on. For this purpose, first, the reviews from known platforms are collected and then they are classified into a few categories like good, average, below average, and bad. At last, the results are shown in terms of an understandable report to the user, helping him/her to get into their dream university. In addition, this paper gives an outlook to further research in the context of an automated analysis of social media content to support the evaluation of universities.

The pedestrian detection algorithm (PDA) is one of the most widely used techniques in modern automated vehicles, surveillance systems, human-machine interfaces, intelligent cameras, robots, etc. Despite considerable work in this field, PDA is still receptive to several scopes of advancements considering some adverse weather conditions like fog, rain, low visibility, etc. Along with this, there are certain intricate scenarios where the accuracy of a given PDA becomes contentious. As we are progressing toward autonomous vehicles, it becomes vital for such vehicles to ensure the safety of both passengers and pedestrians walking around the road. To do so, they require a much more reliable and effective pedestrian detection system capable of working under adverse conditions. This paper considers all such issues to develop certain machine learning (ML) and deep neural network (DNN) methods to solve such issues. YOLOv4 is a deep learning-based object identification method that is currently functioning well yet is not robust. The core premise of YOLOv4 is initially explored and evaluated in this paper to discover its importance in our task. This research devises a coupled system capable of detecting pedestrians under various adverse and intricate scenarios. To do so, we use the YOLOv4 object detection technique coupled with some image denoising, low light enhancement and image dehazing features. We are using the wavelet and YCbCr methods for image denoising and low-light enhancement. To dehaze the video frames, we use airtight estimation and tuning the transmission by deriving the boundary constraints. The paper tries to cover most of the aspects that an autonomous vehicle may face while on the road. Overall, we deliver a reliable model that fosters more accuracy even in complex scenarios and unfavourable weather conditions.

According to the World Health Organisation, falling is a major health problem with potentially fatal implications. Each year, thousands of people die as a result of falls, with seniors making up 80% of these fatalities. The automatic detection of falls may reduce the severity of the consequences. Our study focuses on developing a vision-based fall detection system. Our work proposes a new feature descriptor that results in a new fall detection framework. The body geometry of the subject is analyzed and patterns that help to distinguish falls from non-fall activities are identified in our proposed method. An AlphaPose network is employed to identify 17 keypoints on the human skeleton. Thirteen keypoints are used in our study, and we compute two additional keypoints. These 15 keypoints are divided into five segments, each of which consists of a group of three non-collinear points. These five segments represent the left hand, right hand, left leg, right leg and craniocaudal section. A novel feature descriptor is generated by extracting the distances from the segmented parts, angles within the segmented parts and the angle of inclination for every segmented part. As a result, we may extract three features from each segment, giving us 15 features per frame that preserve spatial information. To capture temporal dynamics, the extracted spatial features are arranged in the temporal sequence. As a result, the feature descriptor in the proposed approach preserves the spatio-temporal dynamics. Thus, a feature descriptor of size (Formula presented.) is formed where m is the number of frames. To recognize fall patterns, machine learning approaches such as decision trees, random forests, and gradient boost are applied to the feature descriptor. Our system was evaluated on the UPfall dataset, which is a benchmark dataset. It has shown very good performance compared to the state-of-the-art approaches.

The Internet nowadays is playing a significant role in our everyday life. Many web services have registration and require user input. These involve completely automated public Turing tests to tell computers and humans apart (CAPTCHA). The web services like Gmail registration, bank transactions, and ticket booking often face attacks from bots which are computer programs. The CAPTCHA is used to protect the sites from these dangerous bots. This technique offers good security for websites as analysis proves that humans can easily solve the CAPTCHA but bots can rarely do that. This is because computers do not have the visual capacity and real understanding like humans. There are four types of methods in the development of CAPTCHA. They are CAPTCHA based on images, audio, video, and puzzles. Here the paper is focused on the development of Puzzle-based CAPTCHA. This implies the user needs to solve the puzzle to access the website. If the puzzle is not solved, the user can't go forward as the site treats it as a bot and denies access. This is one of the efficient methods which made it easily possible for humans but hard for bots. It involves human intelligence and simple testing. Various puzzles are used as a CAPTCHA. This paper is based on the pattern recognition puzzle CAPTCHA. The basic idea is that the user needs to identify the next pattern based on existing patterns. This involves simple logic which humans can solve. The pattern recognition puzzle challenges human intelligence and is much more difficult for the bots to solve so that it could be more secure.

The protection of digital information has drawn significant attention in recent years. Data authentication and copyright protection issues are widely considered as two major concerns. Digital watermarking can be considered as a solution to ensure copyright protection and data authentication in which a watermark will be embedded into a digital content. Later, the watermark can be extracted to ensure copyright protection or to ensure data authentication. Robust watermarking schemes are widely used for copyright protection of digital content and fragile watermarking schemes are preferred for data authentication. The purpose of this work is to provide a thorough analysis of the domain of research regarding image watermarking methods. This manuscript gives an overview of the digital image watermarking approach, the classification of digital image watermarking schemes, and the research challenges in this domain.

Reversible data hiding is an area explored widely in recent days due to its scope of applications in secure message transmission by embedding it in images. The existing RDH schemes in the encrypted image are lagging in terms of embedding rate. In this paper, we propose a new RDH scheme in encrypted images that will ensure the embedding rate without compromising the bit error rate or image recovery. In the proposed scheme, the encrypted image pixels will be classified into black-and-white pixels based on a checkerboard pattern. During the data hiding phase, the sender will select 8 unique random integer sequences S, that form the data keys and whose values are within the range 0 to 255. Data hiding is performed by performing bit-XOR with the white pixels in an image block (size B × B pixels) using one integer sequence from the 8 keys. Each key is correlated to a 3-bit sequence from the secret message. The receiver must have the data keys and the decryption key to extract the hidden message and recover the original image. A smoothness measure between adjacent pixels is defined and used for data extraction and image recovery. The experimental results show that the proposed scheme performs well on the standard image data set (USC-SIPI).

Reversible data hiding (RDH), a promising data-hiding technique, is widely examined in domains such as medical image transmission, satellite image transmission, crime investigation, cloud computing, etc. None of the existing RDH schemes addresses a solution from a real-time aspect. A good compromise between the information embedding rate and computational time makes the scheme suitable for real-time applications. As a solution, we propose a novel RDH scheme that recovers the original image by retaining its quality and extracting the hidden data. Here, the cover image gets encrypted using a stream cipher and is partitioned into non-overlapping blocks. Secret information is inserted into the encrypted blocks of the cover image via a controlled local pixel-swapping approach to achieve a comparatively good payload. The new scheme (Formula presented.) allows the data hider to hide two bits in every encrypted block. The existing reversible data-hiding schemes modify the encrypted image pixels leading to a compromise in image security. However, the proposed work complements the support of encrypted image security by maintaining the same entropy of the encrypted image in spite of hiding the data. Experimental results illustrate the competency of the proposed work accounting for various parameters, including embedding rate and computational time.

As per the details from the World Health Organization (WHO), visual impairment affects 285 million individuals globally. The total number of deaf people is also extreme all over the world. This paper shows the major issues, worries, and obstacles that blind people face, as well as the numerous ways that technology can assist them. This manuscript discusses recent advancements in the domain of assistive technology for blind and deaf people and also aims to give an overview of the assistive technologies available for the blind and deaf so that the researchers can work on this domain to come up with new solutions, or they can improve the existing products. Various research challenges in this domain are also briefly discussed in this manuscript.

Reversible data hiding (RDH) is an active area of research in the field of information security. In RDH, a secret can be embedded inside a cover medium. Unlike other data-hiding schemes, RDH becomes important in applications that demand recovery of the cover without any deformation, along with recovery of the hidden secret. In this paper, a new RDH scheme is proposed for performing reversible data hiding in encrypted images using a Fibonacci transform with an ensemble learning method. In the proposed scheme, the data hider encrypts the original image and performs further data hiding. During data hiding, the encrypted image is partitioned into non-overlapping blocks, with each block considered one-by-one. The selected block undergoes a series of Fibonacci transforms during data hiding. The number of Fibonacci transforms required on a selected block is determined by the integer value that the data hider wants to embed. On the receiver side, message extraction and image restoration are performed with the help of the ensemble learning method. The receiver will try to perform all possible Fibonacci transforms and decrypt the blocks. The recovered block is identified with the help of trained machine-learning models. The novelty of the scheme lies in (1) retaining the encrypted pixel intensities unaltered while hiding the data. Almost every RDH scheme described in the literature alters the encrypted pixel intensities to embed the data, which represents a security concern for the encryption algorithm; (2) Introducing an efficient means of recovery through an ensemble model framework. The majority of votes from the different trained models guarantee the correct recovery of the cover image. The proposed scheme enables reduction in the bit error rate during message extraction and contributes to ensuring the suitability of the scheme in areas such as medical image transmission and cloud computing. The results obtained from experiments undertaken show that the proposed RDH scheme was able to attain an improved payload capacity of (Formula presented.) bits per pixel, outperforming many related RDH schemes with complete reversibility.

-

Among all the approaches, Digital watermarking is the most widely implemented approach for copyright protection and authentication of data. In this technique, a unique piece of information is known as a watermark. Then the watermark gets into an image, later, to achieve its objective the watermark will be extracted. For the transmission of medical images, digital watermarking schemes are mostly used to ensure that the image has not gone through any unauthorized or illegal modifications during the transmission. Since conventional watermarking schemes alter the pixels in the original image, it is not suited for watermarking medical images. In medical images, permanent modifications may adversely affect the diagnosis process at the receiver side, caused by watermarking, especially when we are using some computer-aided diagnosis tools. This motivated computer scientists to work on reversible watermarking schemes. The reversible watermarking technology makes it possible to recover the required medical image from the watermarked image, while extracting the hidden watermark. So, the reversible watermarking technique does not affect the diagnosis in any way since the recovered image will be equivalent to the original image. This recovered image will be used by the user. The use of reversible watermarking techniques to send patient reports along with medical images is also explored, with the patient reports being embedded in the medical picture itself rather than the watermark. These techniques are commonly known as reversible data hiding techniques. This book chapter gives a brief overview of reversible data hiding techniques, reversible watermarking methods, and the major applications in medical image transmissions. In addition, the chapter addresses contemporary reversible data hiding and reversible watermarking algorithms intended specifically for medical picture transmission. The chapter also discusses some of the obstacles that must be overcome when developing a reversible watermarking system for healthcare applications.

Human activity recognition plays a prominent role in applications like sports, violence detection, accident detection, women security, and smart homes by predicting abnormal human behaviour. Human activity recognition now offers a solution for fall detection systems that assist older people at home. A fall detection system that responds promptly to fatal falls can help to reduce the fall severity. Fall detection systems are developed using wearable environmental sensors and video data. Wearable devices are not always feasible as they cause inconvenience to the user. Hence, our paper presents a new promising solution for fall detection using vision-based approaches. In this approach, we analyse the human joint points which are the prime motion indicators. A set of key points of the subject are acquired by applying the AlphaPose pre-trained network. These keypoints are inferred to be the joint points of the subject. The acquired keypoints are processed through a framework of convolutional neural network (CNN) layers. Here, the spatial correlation of the keypoints is analysed. The long-term dependencies are then preserved with the help of long short-term memory (LSTM) architecture. Our system detects five types of falls and six types of daily living activities. We used the UP-FALL detection dataset for validating our fall detection system and achieved commendable results when compared to the state-of-the-art approaches. For comparison, we employed the OpenPose network for keypoint detection. It is inferred from the results that AlphaPose network is more precise in keypoint detection.

A critical issue with the current communication revolution is data security and privacy, which is an inevitable part of trustworthiness in the communication system. Hence, the applicability of the reversible data hiding schemes (RDH) in this scenario is encouraging and critical. Like medical image communication, satellite image transmission, etc. Earlier, we explored Arnold transform in one of our previous works to hide the secret data that uses the convolutional neural network (CNN) model to design a complete RDH scheme. The proposed scheme follows a statistical approach to support recovering the cover image and the embedded information. This approach proves advantageous over the previous work following its computational capability. The scheme designed can retain the entropy of the encrypted images even after embedding the additional information, complementing the security of the encryption algorithm.

Content-based image retrieval (CBIR) is an active area of research due to its wide applications. Most of the existing CBIR schemes are concentrated to do the searching of the images based on the texture, colour, or shape features extracted from the query image. In this manuscript, we propose an object detection-based CBIR scheme with quantised colour histograms. In the proposed scheme, the meaningful objects will be identified from the query image by using you only look once (YOLO) object detection techniques and the quantised histograms of each of the object categories. The object lists, their count, and the area covered by the objects along with quantised colour histograms will be used during feature matching to retrieve the related images from the large image pool. The experiment of the proposed scheme is carried on the Corel 1K and Caltech image dataset. We have observed an average precision of 0.96 during the experimental study which is quite high while comparing the precision from the well-known existing schemes.

Biometric authentication is used in its fullest potential in all smart environments. Face recognition and fingerprint recognition are the two most popular approaches for biometric authentication. In some places, where a higher level of security is required can be equipped with the combination of face recognition and fingerprint recognition. In general, from the client side, the biometric data will be collected using a camera and fingerprint scanner, and that information will be transmitted to a cloud service provider to do the complex tasks, such as face recognition and fingerprint recognition. In such situations, the transmission of face images and fingerprints from the client side to a server is a major concern since an intruder may try to grab that information, and later, they can use that information to control the authentication system. In this chapter, we discuss a secure mechanism to send the face images and the fingerprints with the aid of a reversible data hiding (RDH) through the encryption scheme. RDH is a process of hiding some data by using a concealing medium in such a way that later the extraction of the hidden message is possible along with the recovery of the original images. In a RDH through encryption scheme, the RDH process and the image encryption process will be combined into a single task to hide the secret message in an image. In this chapter, we propose a new model in which the compressed fingerprint data as a secret message will be embedded into the face image through a reversible data hiding through the encryption scheme. The encrypted image obtained after RDH through encryption will be transmitted to the cloud service provider for further processing.

Due to evolution in every sector around us, a large amount of information is generated at every moment. In recent years, many microblogging platforms have become the primary place for the public to express their mood. So the sentiment analysis data collected from websites, reviews, and feedback can be used in many areas such as in marketing or product analysis or competitive research for future development. So, sentiment analysis is one of the booming domains, a study of analyzing people's reviews, suggestions, feedback from the text format and checking if the data shows positive, negative, or neutral emotion. Applications of sentiment analysis are endless and are applicable in every social domain. We can automatically assess the tone of internet interactions using natural language processing and machine learning methods. Sentiment analysis can be performed with the help of various techniques, depending on how much data needs to be examined and how precise the model needs to be. This manuscript is a comprehensive article that includes all the essential topics related to sentiment analysis, such as its applications, challenges faced and various algorithms.

Vehicle detection and number plate recognition approaches have been widely studied in recent years due to their wide applications. In this research paper, we propose a framework to ensure the entry of authorized vehicles in restricted areas such as the University campuses, townships, etc., where we are expecting the entry of a set of authorized vehicles. Certainly, unauthorized vehicles might be allowed to enter those areas after proper verification by the concerned people responsible for ensuring security. In the proposed approach, the admin should register all the authorized vehicles in a system with the essential attributes such as vehicle number, type, etc. A surveillance camera placed at the entrance will capture live videos. When there is a vehicle in the camera view, the image frames will be passed to an automatic number plate recognition module. The number plate recognition module will identify the same, and it will be matched with the details in the database to authorize the vehicle. This manuscript proposes a real-time and reliable approach for detecting and recognizing license plates based on morphology and template matching. To ensure the reliability of the system, a frame selection module will select the image frames with high quality, and even to improve the number plate recognition accuracy, the image will be enhanced using image enhancement techniques such as histogram equalization. The image enhancement techniques will help to provide better results even though the videos are taken with low lighting conditions. Further, we ensure that the vehicle type matches the number present in the database to prevent unauthorized access using fake number plates. The experimental study is conducted using videos taken under various environmental conditions such as lighting, slope, distance, and angle.

Reversible data hiding (RDH) is a recently emerged research domain in the field of information security domain with broad applications in medical images and meta-data handling in the cloud. The amount of data required to handle the healthcare sector has exponentially increased due to the increase in the population. Medical images and various reports such as discharge summaries and diagnosis reports are the most common data in the healthcare sector. The RDH schemes are widely explored to embed the medical reports in the medical image instead of sending them as separate files. The receiver can extract the clinical reports and recover the original medical image for further diagnosis. This manuscript proposes an approach that uses a new lossless compression-based RDH scheme that creates vacant room for data hiding. The proposed scheme uses run-length encoding and a modified Elias gamma encoding scheme on higher-order bit planes for lossless compression. The conventional Elias gamma encoding process is modified in the proposed method to embed some additional data bits during the encoding process itself. The revised approach ensures a high embedding rate and lossless recovery of medical images at the receiver side. The experimental study is conducted on both natural images and medical images. The average embedding rate from the proposed scheme for the medical images is 0.75 bits per pixel. The scheme achieved a 0 bit error rate during image recovery and data extraction. The experimental study shows that the newly introduced scheme performs better when compared with the existing RDH schemes.

Extraction of useful information from a huge volume of data is an active area of research. In this, we have considered a new problem in which we need to extract a set of predefined attribute values from an image pool, where the images in the image pool may belong to a particular person. The objective of this research is to process a set of images of a person to identify the values for various attributes such as age, gender, marital status, hobbies, physical attributes, etc. The major challenge in the proposed scheme is to identify the key face after face extraction from the image pool. The key face image will be the image that will repeat more times and there is a high probability that the image pool belongs to that person. We have used a set of state-of-the-art approaches to identify the attribute values such as age, gender, etc. from the key face image. The ground truth data are collected from them before the validation of the proposed scheme.

Reversible Data Hiding (RDH) is an active area of research having numerous applications in the field of medical image transmission for transmitting clinical reports along with medical images. The existing RDH schemes are categorized into RDH in natural images, RDH in encrypted images, and RDH through encryption. In this research work, we have considered RDH through encryption and used advanced standard encryption techniques. An RDH through encryption scheme in medical image transmission will take a medical image, a clinical report (text file) and an encryption key as inputs and produce an encrypted image as the output. Note that the clinical report is embedded in the medical image as part of the encryption process. In the proposed scheme, the original image is divided into non-overlapping blocks, and AES encryption is performed on each block to get the encrypted image. The key used in AES encryption is scrambled by Arnold transform d times for each block, where d is the data bit that is to be embedded in that block. At the receiver side, the data extraction and image restoration will be carried by analyzing the texture property of image blocks generated through the AES decryption process with all the Arnold transform versions of the AES key. The experimental study is done on medical and natural images to analyze various efficiency parameters such as embedding rate, bit error rate, Peak Signal to Noise Ratio (PSNR), and Structural Similarity Index (SSIM).

A brain tumour is described by the presence of abnormal cells in the brain's tissues. Brain tumours can be benign (not cancerous) or malignant (cancerous). The malignant brain tumour is one of the leading and common cancers in the world. There are two types of tumours, primary tumours that develop in the brain and secondary tumours that start in another region of the body and then spread to the brain. The precise identification of the size and location of a brain tumour is crucial in the diagnosis of a brain tumour and is often diagnosed with magnetic resonance imaging (MRI). This book chapter discusses the major types of brain tumours and the advancements in computeraided approaches for detecting brain tumours. The manuscript gives an overview of various recent machine learning and medical image processing approaches developed recently for the identification and classification of brain tumours. Several medical image dataset available for the research works in this domain is also briefed in this article. The major research challenges which can be addressed by the researchers in the domain of brain tumour detection are also discussed in this article.

Due to COVID-19, the attempt to access medical facilities has increased. But, physical meetings are the biggest reason for the global pandemic. The healthcare industry is in a dire state of despair because of the spread of COVID, constrained health workers, and infrastructure. Healthcare accommodations are costlier than ever due to various reasons such as the global population getting busier and the rise in the number of chronic diseases. Nowadays, there is a high demand for healthcare support due to the increase in the aging population, the COVID-19 crisis, its postdisease difficulties, and other chronic diseases. Treating all the patients at hospitals is not practical nowadays due to limited healthcare facilities. The patients' care is all dependent on visits to the hospital physically, and to some extent, calls, text messages, and emails. This way it was not possible for the doctors to monitor the health of the patients continuously and assist them whenever needed. Recently, the Internet of Things (IoT) and various machine learning approaches have been widely explored in this sector. The IoT-based healthcare methods provide quality and efficiency during treatment. The IoT along with machine learning techniques can transform the healthcare industry by making it more accessible and less costly and make healthcare more facile by equipping the users with pocket amicable medical facilities. With IoT-enabled devices, remote monitoring can enable patients to better connect with doctors, thereby minimizing physical visits. It can equip doctors to easily keep track of their patient's health by utilizing wearables like fitness bands that can monitor various healthcare parameters. All these systems mostly use machine learning or deep learning approaches. This way, they will know whether to offer any immediate medical care or to make any changes to the current treatment. This chapter gives a detailed discussion on IoT and machine learning in the healthcare sector and the security concerns in this domain along with the major research challenges.

A Review of Face Recognition Technology: In the previous few decades, face recognition has become a popular field in computer-based application development This is due to the fact that it is employed in so many different sectors. Face identification via database photographs, real data, captured images, and sensor images is also a difficult task due to the huge variety of faces. The fields of image processing, pattern recognition, and computer vision are all crucial to face recognition. Almost 20 papers from 2011 to 2021 were analyzed and it was observed that a framework called Deep Learning includes a number of significant algorithms. To get better results, you need to employ distinct network models for different applications (images, speech, and text). A review of major person recognition research is also included which is followed by a detailed description of the most accurate and latest datasets (like CKPlus, DeepFace, ImageNet, etc.) required in face recognition. Further, the properties of appropriate face authentication applications and architectures are discussed. This paper also includes the research challenges that might occur and some potential future works as well. We present a prospective analysis of facial recognition also in this manuscript.

The recent advancements in technology widely help to substitute manpower with machines in a better way. Even though machines are increasingly replacing humans in various ways, there are still a few areas where the use of machines is still needs to be explored much more efficiently. Facial recognition systems are one such field. Facial recognition systems are used with various motives, such as identification of suspects in public places, authentication of users on restricted premises, etc. In this work, we propose a facial recognition system to facilitate the authentication of students at the university entrance. The same scheme can be utilized to authenticate the students before entering examination halls also. As the strength of the students at our University (other educational institutes also) increases in a larger way, it becomes strenuous for the security people to record their attendance manually, which frequently results in erroneous data. In this work, we propose a facial recognition system that will help to capture the live videos from an area of interest and identify the faces. Further, a face recognition scheme will detect whether the person is authorized or not. Several facial recognition systems are already available in the literature, and our scheme is different from them in many ways. The proposed method selects the frames with less blur for face detection and further face recognition. A blur detection scheme is used in the proposed system to analyze the amount of blur in the image. To overcome the challenges such as low accuracy during face recognition when the images are taken in low lighting conditions, we use a histogram equalization method to enhance the quality. The experimental study shows that the proposed approach works well in real-time.

Stock market forecasting has been one of the most interesting subjects for many professionals and researchers to work on as a result of today's rapid expansion. Economic conditions, investor sentiment, current events, future guidance, and a variety of other factors all have an impact on the stock market. And because the stock market changes swiftly from time to time, it might be tough for a user or investor to keep up with the shifting trend. Combining sentiment analysis with a machine learning model would help with accurate prediction in this case. Where sentiment analysis is a text mining procedure that has one of the most important uses in analysing user reviews and evaluating the overall sentiment of a piece of text. The purpose of this research work is to create a machine learning model that takes recent tweets from the Twitter API and categorises each message as positive, bad, or neutral. Later, depending on the impact and the person who wrote the tweet, a factor will be multiplied. The proposed method would then take into account the post's owner's total number of followers, as well as the emotion of each comment on each post of selected stock, as well as the number of likes and retweets, utilising market indicators and pricing. The user would be given an overview of the selected stock's potential.

In recent years, reversible data hiding (RDH) in encrypted images got much attention due to its wide applications in the areas such as cloud computing, military image transmission, medical image transmission, etc. This paper introduces a new solution for reversible data hiding in encrypted images. One of the main challenges while designing a reversible data hiding scheme in an encrypted image is the embedding rate and bit error rate during image recovery. The scheme proposed in this manuscript ensures a good embedding rate and the lossless recovery of the original image. The key idea behind the proposed technique is that the encrypted image will be partitioned into non-overlapping blocks, and the pixels in each block will be categorized into white pixels and black pixels based on a predefined pattern. The black pixels will be mapped into a new pixel value based on the two bits from the secret message that is to be embedded into the selected image block. For mapping purposes, we generate four different random permutations of all the possible gray-scale values (0 to 255). At the receiver side, corresponding to each block in the image we have to generate four different weighted mesh graphs. The image recovery and data extraction are carried out by analyzing the total edge weight of these mesh graphs. The results obtained from the experimental study are much better while comparing with a few of the well-known recently introduced reversible data hiding schemes in encrypted images.

The process of coloring neat sketches is a significant activity while making animated movies or for better visualization in computer modelling. The color filling tools are widely available in almost all the image/video editing software, which will help us to pick a color from a color palette, and it can be filled on a selected region. This process is known as flat coloring. The flat coloring process is having a number of challenges. One of the major challenges is that the color may leak from the selected regions to other neighbouring regions if there are some small openings on the contours. The second concern while using flat coloring is that the whole selected regions will be completely filled the same color and hence the drawing will not have an artistic look. In this paper, we propose a new software application that will take a neat sketch as the input and the system will generate a colored drawing as the output. In the proposed scheme, we have converted the given sketch as grayscale or binary image, and we applied image dilation operation to fill the small open spaces in the contours (if any). Further, the closed regions are identified and colored with a predefined set of colors or with random color combinations. While coloring the regions, the proposed system will ensure that the adjacent regions will not colored with the same colors. A number of sketches have been considered during the experimental study and the results are validated manually.

The research in the domain of reversible data hiding (RDH) is recently explored in all aspects due to its applications in cloud computing, forensics, and medical image communication. In this manuscript, we introduce a RDH scheme in encrypted images which can provide a high embedding rate without compromising the bit error rate during the message extraction and image recovery. The proposed scheme follows a block-wise data hiding process. If the processed block size is A×A pixels, and if A=2n+1, then the data hider can hide any number from the set {0, 1, …, (A+2n?1) } in that block. We have introduced an Arnold transform-based data hiding process in which each block will be undergoing a series of scrambling processes based on the bit sequence that we need to embed in the selected block. The message extraction and image restoration are carried out at the receiver side using a trained convolutional neural network (CNN) model.

According to late insights, worldwide, essentially 2.2 billion individuals have some sort of difficulty related to the visual system. The blind people are facing various issues to travel from one place to another, to interact the with the other people in a social event, the blind people may completely fail to understand the emotions of the people with them. The advancement in technology gives various kinds of support for blind people. This manuscript gives a detailed review of various tools and technologies available for assisting blind people. This study deals with the techniques currently under research in this field, the products available in the market which will be useful for blind people, the cost details, and the other advancements in this area. This manuscript will be a useful reference for the researchers who wish to work to design and implement some new products to assist blind people.

-

Road accidents caused due to drowsiness of the driver are quotidian. As per the World Health Organization global report, India has the highest number of road accidents, and about half or greater number are because of drowsy driving, and this has become a major issue. Real-time drowsiness detection models detect when the driver is feeling drowsy by monitoring behavioural aspects or by using physiological sensors. Though the use of bio-sensors gives more accurate results, they are intrusive and distract the driver. We have developed and implemented a behavioural-based drowsiness detection algorithm that monitors the movement of the face and closeness of eyes to detect and alert a drowsy driver. We successfully implemented our algorithm in Matlab-2020 software, where we took a live video from a webcam and processed each frame to classify it as either drowsy or not. We also tested on a dataset featuring live driving subjects and achieved 90% accuracy with 84% precision. If drowsiness is detected, a system audio alert is generated to alert the driver. In case eyes or face are not detected in a frame, we by default classified it as drowsy and produced the alert message because a false negative is more dangerous than a false positive, and thus attained a high recall of 98%.

Open-Source Software has picked up pace in the past decade with support from Multinational conglomerates and huge Open-Source communities. We hear a lot about the success of many open-source projects, but we fail to understand how many do not make it. In this paper, we understand the dynamics behind open-source software. We start with the need for Open-Source Alternatives. Then look at a few concerns faced by Open-Source Software developers and maintainers. Next, we would understand the various requirements of Open-Source Software. Later, we would touch upon the various attributes that affect the selection of Open-Source Software and the decisions to be taken while building general-purpose Open-Source Software. Then we would analyze the 5-determinants of Open-Source Software success. Finally, we would look at the data collected from 482 datapoints from 24 countries and then analyze the data by forming graphs and charts.

Technological advancements and digitalization have made the life of humankind simple but at the same time imposing many challenges. As information started bursting across the internet, information management and security became major concerns. Recently, researchers have been focusing on a hot topic called reversible data hiding (RDH). RHD secures the data by covering it within another medium. It allows the recovery of the medium and hidden information on the receiver side without any loss. This work discloses a high capacity RDH scheme in the encrypted image with a Fibonacci transform image scrambling algorithm for data hiding and a convolutional neural network (CNN) based recovery. It follows a block-wise embedding process, embedding (n + 1) bits within a block of size 2n while n > 1. The proposed scheme is tested on the USC-SIPI image data set from the University of Southern California and has resulted in an improved embedding rate compared to the existing Arnold transform-based RDH and many other well-acknowledged RDH schemes.

-

Reversible data hiding (RDH) is a widely studied research domain due to the extensive applications in medical image transmission, military image transmission, and cloud computing. In this paper, we propose a novel RDH scheme in encrypted images which is having a high embedding rate without affecting the bit error rate. To perform data hiding, all the non-overlapping pixel sub-arrays from the encrypted image will be considered in a predefined order, and one-bit from the secret message can be hidden in each of the pixel sub-arrays through a controlled pixel mapping technique. The pixels in a selected sub-array P will be split into two groups: odd pixels and even pixels according to the parity of the index of the pixels. All the even pixels will be replaced by a new pixel value based on a predefined mapping function to hide a secret bit 1. To embed bit-value 0, no need to do any modifications in the pixel sub-array P. A new correlation measure between adjacent pixels is introduced in this manuscript to extract the secret message bit and to restore the original image. All the simulations of the proposed scheme are done on the standard images from USC-SIPI image dataset. The results show that the proposed scheme achieved a very good embedding rate without affecting the bit error rate or image recoverability.

Video search is an active area of research in the field of information retrieval. The major approaches explored for video retrieval are metadata-based search and content-based video retrieval. In this manuscript, we propose a video retrieval system that will retrieve the videos based on the textual information present in the video frames. The key idea behind the proposed scheme is that the user who needs to retrieve a set of videos will give a keyword text through user consol. The video retrieval system will extract the frames from all the videos in the video pool and an optical character recognition module extracts the text information from each of the frames. The presence of search keywords in each frame will be analysed using the pattern matching technique. The number of times a search keyword is present in a video will decide the rank of that particular video in the final search results. To reduce the time requirement of the video retrieval operation, we have considered one frame that belongs to a scene. A video may consist of several scenes, some of the absolute differences between the reference frame and the frames video will determine the scene changes. The experimental study of the proposed scheme is carried out on the videos downloaded from National Programme on Technology Enhanced Learning (NPTEL) website. The proposed scheme will be useful to academicians to search the videos of a particular topic. The results on a limited video dataset show that the proposed scheme is performing well in real-time situations.

A Reversible data hiding through encryption (RDTE) scheme will consider an original image and a sequence of bits as the input and generate an encrypted image as the output. This encrypted image will be able to transmit through the network securely, and the authorized receiver can take out the hidden details along with the restoration of the actual image. This paper proposes a new RDTE scheme with a good rate of embedding without any issues during the restoration of the original image. We have used the well-known RC4 pseudo-random generator for the image encryption, and we performed data hiding during block-wise image encryption. In the proposed scheme, the original image is considered as non-overlapping blocks of size B × B pixels, and these blocks will be encrypted using a sequence of pseudo-random integers. During the RDTE process, all the possible unique permutations of the encryption key, K, will be generated, say (K0, K1,.., KN). Further, the sender will be capable of embedding one integer value from the set {0, 1,.., N} in a selected image block. A selected block will be encrypted using the pseudo-random sequence of integers using the key KQ to embed the integer Q in the selected block. The proposed scheme prefers to select keys with unique characters with sufficient length to ensure the maximum embedding capacity. The message extraction and image restoration are performed by analyzing the entropy measure from each block after attempting the decryption.

The global rise in population has led to a shortage of raw materials and food supplies. The agricultural sector has become the primary and most vital source to overcome this particular constraint. However, the industry itself is facing the challenge of pests and various crop diseases. Battling this has been the significant focus of the sector for decades. Still, due to the technology gap that existed earlier, there existed a constraint on identifying the diseased crops on a massive scale. Nevertheless, today, with the improvement of technologies such as drones, IoT devices, and higher processing speeds combined with data analysis and machine learning, the problem of identification can be resolved quickly. This paper aims to provide a brief description of existing solutions that have been published and focuses on the more efficient machine learning model based on conventional neural networks (CNN) that we have developed. This machine learning model can be deployed on IoT devices, mobile phones, and drones and cameras that farmers can utilize to identify the diseased crops on a massive scale and take the necessary precautions not to let the disease spread and affect the supply produced. The proposed model using CNN was trained using images from plant village dataset and attained an accuracy of 94.87% in identifying the diseased plants with the help of image processing by OpenCV. Finally, the paper showcases the detailed analysis of the proposed scheme and results obtained by the model.

Reversible watermarking is an active research area in the field of data security. In this manuscript, a reversible watermarking scheme able for electronic patient record (EPR) transmission and medical image authentication is introduced. The proposed scheme computes an adaptive authentication code from the image to be watermarked. Further, the watermark (authentication code and EPR) are embedded into the medical image through a novel image scaling up operation. While scaling up an original image with M×M pixels by a factor of 2, 3M2 number of pixel intensity values in the scaled-up image needs to be approximated by considering the original pixel values. In this work, a set of rules is defined to insert authentication code and EPR data through the copying of neighborhood pixels as the missing pixel intensity values in the new scaled-up image. When the neighborhood pixels have the same intensity value, the adaptive pixel copy operation will not be successful to insert the watermark. In such cases, the bit in the least significant position uses for the watermarking process. The standard medical images downloaded OsriX dataset is used for experimental study. The results obtained show that the new scheme introduced in this manuscript surpasses the existing reversible watermarking. The scheme introduced in this paper can be used as part of an automated system where EPR data needs to be transmitted along with the medical images for better healthcare services.

Many countries including India are frequently affected by the natural disasters like floods. In general, predicting natural disasters accurately is very difficult, but advanced technologies can be utilized to come out of such difficulties or to reduce the impact of natural disasters. Satellite image processing is one of the efficient ways to detect water bodies in earth regions which may help the agriculture industry or to identify the flooded regions. In this paper, we propose a scheme to identify the water bodies from the satellite images which will be useful for various applications. During our research, we have created a set of water body images by cropping satellite images. The properties of the water body regions analyzed using an algorithm and computed a set of possible threshold values for the pixels representing the water bodies. The threshold values obtained from the analysis of 'water body images' are used in the proposed algorithm to identify water bodies in any given image. A sequence of morphological operations is introduced to refine the results that are obtained through pixel color component analysis. The result analysis is carried out on a set of satellite images and it achieved good results.

Human activity detection is an active research topic now, the difficult problem of fine-grained activity detection is often ignored. This paper proposes a method to detect human activity from still images. Iterative detection of human activity in a scene is another tough and exciting area of computer vision research. In our day to day life, we have seen implementations of automated cars, speech recognition, and various machine learning models. Unlike action detection in videos that have spatio-temporal features, still images can't be considered similarly, making the problem more complex. The current work solely comprises activities that involve objects to reach a simple answer. Based on semantics, a complicated human activity is broken down into smaller components. The significance of each of these elements in action recognition is investigated in depth. This system is based on detecting an individual's action or behaviour with the help of a single frame (image). Activity detection consists of various tasks like object recognition, pose estimation, video action recognition, and image recognition. Since the current paper is focused only on actions that involve objects, a dataset with specified classes is created. Images for this dataset will be chosen from different sources. This study aims at the development of computational algorithms for activity detection in still images.

Reversible data hiding (RDH) in encrypted images is an emerging research domain in the field of data security. Since the embedding rate from the existing RDH schemes in encrypted images is low, those schemes are not well suited for various real-life applications such as medical image transmission and cloud computing. To resolve this issue, in this manuscript we introduce an RDH scheme capable of providing high embedding without compromising the other efficiency parameters such as bit error rate and image recoverability. The proposed scheme uses a block-wise data hiding process in which a block of size B×B will be considered from the encrypted image at a time, and that will be scrambled through a sequence of Arnold transform. The bit sequence that the data hider wants to hide in a block will decide the number of Arnold transforming operations on the block. The same process will be continued for all the blocks in the image to get a final encrypted image with a hidden message. At the receiver side, the extraction of the hidden message and the image recovery is carried out with a trained support vector machine (SVM) model. The SVM model is capable to predict a given image block into any one of the two classes: encrypted block or natural block. The experimental study of the proposed scheme is carried out in the USC-SIPI image dataset and the results show that the new scheme surpasses the recent well-known RDH schemes.

Reversible data hiding (RDH) is an actively emerging area in the domain of Information Security having wide applications in clinical data transmission along with medical images. In our research, we came up with a new RDH scheme to keep clinical data in the medical image to ensure secure data transmission. Histogram shifting-based RDH schemes are widely explored for RDH in images. The conventional histogram shifting-based RDH schemes have two major concerns: low embedding rate and overflow or underflow. In this approach, we discuss a prediction error histogram shifting-based approach with an improved overflow handling technique. The pixels in the images are divided into two different categories: black and white. The classification of the pixels has been carried out based on the checkerboard pattern. As we know that as per the checkerboard pattern, a black pixel will have four 4-neighbourhood pixels (left, right, top and bottom). To predict the black pixel value in the middle we used the average of three pixels out of 4-neighbourhood which are very close to the central pixel value. By considering the predicted pixel value and the actual pixel value, we have computed the prediction error. The histogram of prediction error is generated based on the prediction error corresponds to all the black pixels in the image. The prediction error histogram is considered for further data hiding through the histogram shifting approach. The overflow/underflow is a critical issue in the histogram shifting-based RDH scheme, so we have came up with an improved overflow/underflow handling technique in this approach. We have validated the results after carrying out the proposed scheme on medical and natural images.

Content-based image retrieval (CBIR) is an active research field in the domain of information retrieval. This domain is widely studied in the last two decades and there are lots of approaches are introduced to handle this. Most of the existing CBIR schemes consider shape, color, or texture features of the objects in the image to find similar images for the retrieval purpose. In this manuscript, we introduce a CBIR scheme to retrieve the relevant images based on the textual information present in the images rather than other features. The proposed scheme will take an image as the input and retrieve the images which contain the common text information. From both the query image and from all the images in the image pool, the proposed scheme will identify the text by using the text detection approach. It might be noted that the text detection from natural image scenes is not always working perfectly. So to do the matching between the query image and the images in the image pool, we have used the length of the longest common sub-sequences as the criteria. The experimental study of the proposed scheme is carried and the results show that the proposed scheme works well.

One of the universally studied research topics is a reversible data hiding approach because of its wide applications in medical image transmission, cloud computing as well as in military services. In the introduced scheme stated in this paper, a new RDH scheme in encrypted images is proposed which ensures a good embedding rate without compromising the bit error rate. The scheme introduced in this manuscript uses a new measure that is computationally efficient and ensures real-time performance. In the proposed scheme, the original image is divided into blocks and in each block, one secret message bit is hidden by the data hider. All the pixels in the image block are categorized into two classes: black pixels and white pixels based on the predefined checkerboard pattern. The data hiding is performed by mapping the pixel values in the white position to a new value, only if the data hider wants to hide a secret bit 1. At the time of data extraction and image recovery, the closeness between the adjacent pixels is analyzed by considering a pair of adjacent black pixels and white pixels from the quantized image blocks. The overall experimental study of the proposed scheme is carried out using the standard images from the USC-SIPI image dataset managed by the University of Southern California.

The use of digital images is exponentially increased due to the availability of various image capturing devices and the provision of transmitting them with less expense. Digital images are used for various purposes such as personal, entertainment, medical diagnosis, fashion design, forensic analysis, etc. The retrieval of relevant images from a huge volume of images is a tremendous task. Content-based image retrieval (CBIR) is a solution to retrieve relevant images from a pool of images by considering a query image. In a CBIR scheme, the user will give a query image and the system supposes to return a set of the image which are very similar to the query image. The CBIR schemes use classification models and the visual features of the images while ranking and retrieving the images from the image pool. This manuscript discusses the overview of a CBIR scheme, its application domains, the existing approaches in this domain and various datasets available for the research in this domain. The manuscript has also listed a set of challenges in this domain so that future works can be focused on this area.

Due to the coronavirus disease (COVID-19) pandemic, most of the public work is carrying out online. Universities all around the globe moved to online education, job interviews are mainly conducting online, many first-level health consultations are happening online, and companies hold periodic meetings entirely online. Google Meet, Microsoft Team, and other online meeting software applications are widely accessible on the market. In this work, we are addressing a topic that has a lot of practical applications. In this paper, we present a method that takes a recorded video as an input and generates a written and/or audio summary of the same as an output. The suggested method can also be used to generate lecture notes from lecture videos, meeting minutes, subtitles, or storyline production from entertainment videos, among several other things. The suggested system takes the video's audio track, which is then transformed to text. In addition, we created the text summary utilising text summarising algorithms. The system's users have the option of using the text summary or creating an audio output that matches the text summary. The proposed method is implemented in Python, and the proposed scheme is evaluated using short videos acquired from YouTube. Since there is no benchmark measure for evaluating the efficiency and there is no specific dataset available for the relevant study, the proposed method is manually validated on the downloaded video set.

-

-

Histogram shifting-based Reversible Data Hiding (RDH) is a well-explored information security domain for secure message transmission. In this paper, we propose a novel RDH scheme that considers the block-wise histograms of the image. Most of the existing histogram shifting techniques will have additional overhead information to recover the overflow and/or the underflow pixels. In the new scheme, the meta-data that is required for a block is embedded within the same block in such a way that the receiver can perform image recovery and data extraction. As per the proposed data hiding process, all the blocks need not be used for data hiding, so we have used marker information to distinguish between the blocks which are used to hide data and the blocks which are not used for data hiding. Since the marker information needs to be embedded within the image, we have compressed the marker information using run-length encoding. The run-length encoded sequence is represented by an Elias gamma encoding procedure. The compression on the marker information ensures a better Embedding Rate (ER) for the proposed scheme. The proposed RDH scheme will be useful for secure message transmission also where we are also concerned about the restoration of the cover image. The proposed scheme's experimental analysis is conducted on the USC-SIPI image dataset maintained by the University of Southern California, and the results show that the proposed scheme performs better than the existing schemes.

Research into the domain of reversible data-hiding has received a great deal of attention in recent years due to its wide applications in medical image transmission and cloud computing. Reversible data-hiding during image encryption is a recently emerged framework for hiding secret data in an image during the image-encryption process. In this manuscript, we propose a new reversible data-hiding-through-encryption scheme that will ensure a high embedding rate without bringing any additional overhead of key handling. The proposed algorithm can use any secure symmetric encryption scheme, and the encryption and/or decryption key should be shared with the receiver for data extraction and image recovery. As per the proposed scheme, the data hider can hide threebits of a secret message in an image block of a size of B × B pixels. The data extraction and image recovery will be carried out by analyzing the closeness between adjacent pixels. The simulation of the new scheme carried out on the USC-SIPI dataset shows that the proposed scheme outperforms the well-known existing schemes in terms of embedding rates and bit error rates.

Heart disease is one of the most common and serious health issues in all the age groups. The food habits, mental stress, smoking, etc. are a few reasons for heart diseases. Diagnosing heart issues at an early stage is very much important to take proper treatment. The treatment of heart disease at the later stage is very expensive and risky. In this chapter, the authors discuss machine learning approaches to predict heart disease from a set of health parameters collected from a person. The heart disease dataset from the UCI machine learning repository is used for the study. This chapter discusses the heart disease prediction capability of four well-known machine learning approaches: naive Bayes classifier, KNN classifier, decision tree classifier, random forest classifier.

Reversible data hiding (RDH) is an emerging area in the field of information security. The RDH schemes are widely explored in the field of cloud computing for data authentication and in medical image transmission for clinical data transmission along with medical images. The RDH schemes allow the data hider to embed sensitive information in digital content in such a way that later it can be extracted while recovering the original image. In this research, we explored the use of the RDH through the encryption scheme in a biometric authentication system. The internet of things (IoT) enabled biometric authentication systems are very common nowadays. In general, in biometric authentication, computationally complex tasks such as feature extraction and feature matching will be performed in a cloud server. The user-side devices will capture biometric data such as the face, fingerprint, or iris and it will be directly communicated to the cloud server for further processing. Since the confidentiality of biometric data needs to be maintained during the transmission, the original biometric data will be encrypted using any one of the data encryption techniques. In this manuscript, we propose the use of RDH through encryption approach to transmit two different biometric data as a single file without compromising confidentiality. The proposed scheme will ensure the integrity of the biometric data during transmission. For data hiding purposes, we have used a block-wise RDH through encryption scheme. The experimental study of the proposed scheme is carried out by embedding fingerprint data in the face images. The validation of the proposed scheme is carried out by extracting the fingerprint details from the face images during image decryption. The scheme ensures the exact recovery of face image images and fingerprint data at the receiver site.

Reversible data hiding (RDH) is a widely studied domain due to its applications in cloud computing and in the transmission of the medical images. A new block-wise RDH scheme in encrypted images that uses a checkerboard patternbased pixel selection and pixel inversion technique is introduced in this paper. The use of a checkerboard pattern to classify the pixels into two classes (black pixels and white pixels) and the inversion process on the pixels during data hiding ensures the exact recovery of the image in most of the cases. The extraction of the secret data and the original image recovery is possible by analyzing the correlation between pixels within the decrypted image blocks. A new smoothness measure is introduced in this paper which considers all the 3 × 3 pixels in the overlapping windows of an image block. The new measure introduced in this paper along with the checkerboard pattern-based pixel inversion ensures a low bit error rate. The results claim that the new scheme performs well as compared to the existing techniques. Index Terms-Reversible data hiding, Image encryption, Image recovery, Smoothness measure.

Reversible data hiding through encryption is a recently introduced technique in the area of information security. In a reversible data hiding through encryption scheme, both the image encryption process and data hiding task are combined into a single process. In this paper, we improve the existing reversible data hiding through encryption scheme by introducing a block prechecking mechanism to reduce the bit error rate. The key idea of the proposed scheme is that during image encryption, the original image will be divided into non-overlapping blocks of size n × n pixels and in each block, we can embed three-bit from the secret message. One specific pseudo-random byte stream will be used to encrypt a given image block based on the three-bit data that we want to embed into it. The experimental study of the proposed scheme on the standard medical images downloaded from OsriX data set shows that the proposed scheme outperforms the existing schemes in terms of bit error rate without compromising the embedding rate.

Reversible data hiding on encrypted images is an active area of research in the field of information security. In this paper, we revisit the reversible data hiding scheme proposed by Xinpeng Zhang in 2011. To ensure the exact image recovery at the receiver side, the existing scheme uses a block size of 32 × 32 pixels, this leads to an embedding rate of 9.76 × 10 bite per pixels. We also observed that the existing scheme is prone to fail while recovering an image block which contains highly correlated pixel values (very smooth region). In this manuscript, we propose a new scheme which will help to reduce the block size without compromising the bit error rate. The experimental study of the new scheme is carried on USC-SIPI image dataset managed by the University of Southern California, and the results show that the proposed scheme outperforms the existing scheme.

Automation is a process of utilizing technology to reduce human efforts. In this paper, we propose a computer vision-based automated system for stock management in the supermarkets. The proposed system will help to reduce the manpower required in a supermarket by continuously monitoring the availability product in the supermarket and reporting the useful information to the concerned person automatically. The key idea behind the proposed scheme is that a few low-cost cameras will be placed in the supermarket which will help to capture the videos of the product racks in the supermarket. The presence of human beings are identified by using a structural similarity index (SSIM) based scene change detection technique, further, an object detection technique will be used to count the number of items present in the specific p roduct r ack. If t he number of items present in a particular rack goes below a threshold limit, a short message service (SMS) and/or email will go the concerned authority. To make it more comfortable, a product identifier ( printed) w ill b e k ept j ust b elow t he p roduct racks. An optical character recognition module in the proposed scheme will identify the product identifier a nd i t w ill b e m entioned in the SMS or email which will help the supervisor for scheduling the replacement of the items in the racks. The experimental study is carried out by placing sample items on a rack and the mobile camera is used as an IP camera with the help IP webcam android application for the monitoring purpose. The experimental study shows that the proposed scheme will work reliably in a supermarket environment.

Reversible data hiding (RDH) is an active area of research in the field of information security. For the last few decades variety of approaches were introduced for the purpose of reversible data hiding in digital images. In general, the reversible data hiding techniques can be broadly classified into three categories: histogram shifting based RDH, lossless compression based RDH and difference expansion based RDH. There are many algorithms available in the literature based on the histogram shifting approach. One of the major issues in the histogram shifting based approach is overflow or underflow during the histogram shifting process. In this paper, we propose an efficient technique to handle the overflow or underflow by embedding some additional marking bits in the image during the embedding of the secret message bits. The proposed scheme ensures the exact recovery of the original image and it helps to reduce the overhead of overflow or underflow handling in histogram shifting based RDH scheme. All the experimental study is carried on the standard image data set (USC-SIPI image data set).

Reversible data hiding is widely studied in recent years due to its wide applications in various domains such as medical image transmission and cloud computing. In this manuscript, we propose a novel scheme for performing reversible data hiding in encrypted images. In this scheme, the data hider can embed one-bit additional data bit in a small block (B× 2 pixels) from the encrypted image. All the non-overlapping regions (blocks) in the encrypted image will be processed by accessing those blocks in a predefined order. To embed a bit value 0 in an encrypted image block, no need to modify any pixel values. If we want to embed bit value 1, all the pixels in the first column of the selected image block will be mapped into a new pixel value based on a predefined function. At the receiver side, the data extraction and image recovery are carried by comparing the closeness between pixels in the adjacent columns of the pixels in each block of the decrypted image.

Underwater image enhancement is an active area of research due to its wide applications in areas like marine research, automated underwater vehicles, etc. In general, the underwater images have low contrast, blurriness, and color cast due to various effects like absorption, scattering, and refraction. Normally, the underwater images are less clear and not suitable for various applications. The underwater images should be enhanced to use for real-life applications and the natural image enhancement techniques may not work well for underwater images. In this paper, we introduce a scheme for enhancing the quality of the underwater images by using a residual neural network (ResNet). The synthetic underwater images for the experimental study are generated using the underwater generative adversarial network (UWGAN). The experimental results show that the new underwater image enhancement techniques perform well, and it can be used for real-life applications where we need good quality underwater images.

The design and development of reversible data hiding algorithms got a lot of consideration in recent years because of its wide applicability in medical image transmission and cloud computing. In this paper, we propose a new reversible data hiding scheme though a block-wise multi-level image encryption process to hide lengthy secret messages in an image. In the proposed technique, the data hider will use two additional encryption keys K0 and K1 for reversible data hiding purpose other than the image encryption key K. During the process of data hiding, an encrypted image will be taken as the input and produces a final modified encrypted image with hidden data. For the extraction of data and recovery of image, the receiver needs the image decryption key K and the additional decryption keys K0 and K1. The naturalness property of the image blocks is analyzed to recover the image blocks through a multi-level decryption process. The standard images from the USC-SIPI dataset controlled by the University of Southern California are used to conduct the experimental study.

Design and development of reversible data hiding schemes are widely studied topic due to its wide scope in cloud computing and medical image transmission. This paper introduces a new reversible data hiding algorithm based on the histogram of the blocks of the cover images with an efficient overflow management technique. In the new scheme, the peak intensity value from each block is used for data hiding, and to make sure the correct recovery of the original image, the grayscale value used for data hiding from each block is embedded in the same block itself by replacing the least significant bits of eight selected pixels. The lossless recovery is ensured by embedding those least significant bits in the same block itself along with the secret message. Detailed theoretical analysis and experimental study of the scheme are carried out and discussed in this paper. The images from the standard image dataset of the University of Southern California (USC-SIPI) are used in our study.

-

-

-

Steganography is an active research area in the field of information security. Audio steganography refers to the process of embedding a secret message into an audio signal for secure message transmission. In this paper, we propose a novel secure and robust audio steganography scheme with good embedding rate. The key idea behind the proposed scheme is that a secure encryption scheme is used in the steganography scheme to encrypt the secret message. Further, a random least significant bit plane will be selected by using the genetic algorithm. The encrypted secret message bits will be embedded into the selected bit-plane. The genetic algorithm helps to reduce the distortions on the stego audio after the data hiding process. The selection of higher bit-planes for data hiding process will help to achieve better robustness against noises. The experimental study of the proposed scheme on a variety of audio signals shows that the proposed scheme performs better than the well-known least significant bit steganoaraphy scheme.