Pages: 19

During the COVID-19 pandemic, transmission and sharing of electronic health records (EPR) was encouraged for better diagnosis and treatment accuracy. However, the issues of data exposure, ownership conflicts, and security of these records have attracted more attention from the healthcare research community. To provide better diagnosis accuracy, this chapter presents fusion-based robust and secure watermarking techniques suitable for healthcare applications. Initially, CT scan and MRI images are fused using Non-Subsampled Contourlet Transform (NSCT) based fusion to generate the final fused watermark for better diagnosis accuracy. Then, the combination of NSCT, Hessenberg Decomposition, and Multiresolution Singular Value Decomposition (MSVD) is employed to conceal the fused watermark into the carrier media. Finally, upgraded stereo image encryption is utilized to encrypt the marked image, which is useful for enhancing the security of the media data. The experimental and comparative analysis indicate improved robustness, visual quality, and security of the proposed watermarking technique, with high suitability for healthcare applications.

Pages: 23

The hash functions have been at the center of the data security algorithms for the last two decades. However, with the emergence of quantum computing, conventional methods of data security are under review and scrutiny. At this juncture, the concept of quantum hash function comes to the rescue with its immensely important uses in the field of post quantum data security, especially in quantum key distribution. In this chapter we will present the theory behind quantum hash functions, followed by its uses for post quantum cryptography models, particularly for image data. The algorithms using quantum hash function are also discussed, along with their individual performance.

Pages: 33

The advent of potential scalable and reliable quantum computing poses a serious threat to traditional integer factorization-based security protocols. An entirely new field of post-quantum cryptography has been developed to account for challenges of quantum computing that includes new approaches and algorithms to avoid susceptibility from quantum computers of sufficient capacity. As a sufficiently powerful quantum computer can put up a serious challenge to conventional encryption methods based on the difficulty of integer factorization or discrete logarithms, post-quantum cryptography involves several novel approaches such as Lattice-based cryptography, Multivariate cryptography, Hash-based cryptography, Code-based cryptography, Supersingular elliptic curve isogeny cryptography, and Symmetric key quantum resistance. In this chapter we present a wide survey of post-quantum methods as applied to image security. As the challenges offered by the post-quantum cryptography are an opportunity for the code breakers, the survey presented here includes an appraisal of the challenges and the opportunities offered by the emerging field of post-quantum cryptography from the perspective of image security. Current trends and future directions will also be referenced.

Pages: 17

Biometric authentication is becoming increasingly popular, which offers potential benefits over other traditional authentication methods such as personal identification number, security question, or password. Security and recognition accuracy are the main features of the biometric-based system. Measurement of the biological or physical properties of a human body part, such as a finger, face, hand, voice, retina, iris, etc., can be referred to as biometrics. Moreover, there is no need to remember the password, answer, etc., for the authorization. As time passed, biometric became 3D. Biometric authentication can be made more robust by adding more associated features. With the low cost of the associated sensor, it makes an optimum choice for the security system. This chapter throws light on 3D biometric applications such as personal authorization and personal healthcare, with the advantages and disadvantages of each.

Pages: 23

The proposed method focuses on utilizing the image watermarking technique to improve the security of medical data. Today, the volume of medical data is drastically increasing due to the unavoidable pandemic situation. The proposed scheme uses the properties of lifting wavelet transform (LWT), QR decomposition, Hénon map (HM), and chaotic encryption. The first watermark is the image of the cameraman, and the second watermark contains the patient’s Hénon map information. BRISK (Binary robust invariant scalable keypoints) features are used for authentication. The DICOM (Digital Imaging and Communications in Medicine) ultrasound image of liver is used as the host image. The input image is passed through the LWT to split into sub-bands. Two different sub-bands are used for dual watermarking. The diagonal components of the upper triangular matrix of QR decomposition are used for embedding. The watermarked image is encrypted through chaotic encryption. Peak signal to noise ratio (PSNR), Normalized correlation coefficient (NCC), Kullback-Leibler (KL), and Jensen-Shannon (JS) distances are used as the performance parameters. The results of the proposed technique shows that it has higher robustness and imperceptibility, and also that authentication using the BRISK feature is done successfully.

Pages: 17

Recently, the use of multimedia data has overcome the communication problem among several users. With this data sharing, users can easily exchange their data in a shorter time. But, this sharing of data, such as digital images, also raises the concern of protecting them from intruders, which may tamper confidential information. Therefore, this chapter proposes a novel digital image watermarking technique that utilizes attributes of homomorphic transform (HT) and multiresolution singular value decomposition (MSVD) in wavelet domain. In this algorithm, a cover image is decomposed into illumination and reflectance components with the aid of HT. The reflectance component is being further transformed using redundant discrete wavelet transform (RDWT) and MSVD. Similarly, a watermark image, which consists of a logo, is also transformed using RDWT and MSVD. Embedding a watermark in the reflectance component facilitates better imperceptibility, whereas using RDWT-MSVD enhances the technique robustness against malicious attacks. Pervasive analysis of simulation results demonstrates the efficacy of proposed technique in terms of robustness and imperceptibility, which outperforms other prevailing methods.

Pages: 26

An enormous increase in the volume of multimedia data has been observed, floating around over the internet in recent times. Production of such massive data is mainly due to the easy access to mobile and usage of other high-end Internet of Things (IoT) devices for different services by a large section of mobile users. Services such as video streaming and other multimedia generate petabytes and terabytes of data to the cloud and are bandwidth-intensive. These services demand managing and processing voluminous data, fast reaction time, and improved quality of service with low network latency in real-time, which is a critical task. Good quality of multimedia services is available with cloud computing, but this comes with long processing times. There are also newer technologies like fog and edge computing, which offers storage and managing of resources at a network edge closest to the users, achieving lower latency than the cloud, which maintains sustainable computing and ensures privacy protection. The potential, features, and certain recent research challenges of edge computing are highlighted in this chapter to develop a scalable architecture for managing and processing multimedia data. This chapter also presents a review of proposed edge architectures including Mobile Edge Computing, which aims to create a latency efficient mobile computing platform. Machine Learning, Deep Learning, and Peer-to-Peer networking are also discussed as a solution for applying data analytics and distributed file sharing at edges. The chapter explores the overall concept of edge computing to enthusiastic researchers working with challenges in this area and designing architecture that manages multimedia content at edges.

Pages: 19

Various multimedia forgery such as deepfakes’ detection techniques use deep learning techniques nowadays. But since these deep learning techniques are based on neural networks, whose results are tough to explain because of their inherent black box nature, there is a need to focus on that aspect of deep learning-based detection approaches. Any forensic procedure must have a human-understandable explanation for real-world scenarios. That’s why deepfake detection approaches should be explainable. There are various ways to incorporate explainability in deepfake detection approaches, one of which is Local Interpretable Model-Agnostic Explanations (LIME). This chapter provides a comprehensive overview of explainability in deepfake detection. Various deep learning models are first trained on the Deepfake Detection Challenge (DFDC) dataset, and then evaluation is done based on accuracy metric for each. After that, the LIME tool is applied to justify the classification done by the models.

Pages: 22

The concept of cyber threat intelligence (CTI) is an evidence-based cyber forensic approach that proactively responds to sophisticated security threats through monitoring and evaluating different risk indicators, along with exchanging that information among different industries. Since the method, procedure, and intensity of advanced cyberattacks continue to be progressive, traditional heuristic and signature-based security approaches are unable to prevent these new-age complex cyber threats. Hence, organizations must collect and exchange real-time cyber threat data and convert it into threat intelligence to combat the attacks or accomplish prompt damage recovery. The primary cause of leveraging CTI is to resolve the increasing asymmetry between the cyber-offensive capability of cyber-criminals and the cyber-defensive capability of commercial organizations. Although companies and corporations are attempting to do whatever they can to fight cyber theft, not enough is known about the real threat of CTI. This chapter aims to classify existing advanced threat intelligence types, CTI issues, trends, standards, and emerging research. The chapter also describes the trust and anonymity-based intercommunication strategies for threat information sharing, thus collaborating enterprises can minimize damage caused by business data breaches. This research provides a comparative analysis of the model CTI platforms for sharing the threat detection knowledge to improve the quality of threat intelligence and refine the predictive accuracy to provide a better automated analytic security solution.

Pages: 25

Due to the rapid development of multimedia technology, securing digital images for transmission and storage in cyberspace is a prominent issue. Most existing techniques concentrate on data hiding using blockchain (BC) techniques to address this problem. This chapter summarizes the background knowledge and recent developments in BC watermarking techniques for image security, then categorizes the techniques based on objective, methodology, watermark size, BC’s role in watermarking schemes, evaluation metrics, existing technique limitations, and datasets used for the experiments. Finally, we provide a tabular comparison of state-of-the-art research and discuss potential issues and future directions in this field.

Pages: 22

Communication technology and information systems play a vital role in medical science. Telemedicine applications use communication and information technology to provide healthcare services and enable transferring patient information between distance locations. Medical image security and authorization is an essential component of telemedicine application that ensures the privacy and confidentiality of patient information. During acquisition, coding, authorization, and the encryption process, the image quality deteriorates. Image quality assessment is an important aspect of medical imaging to ensure image quality for accurate disease diagnosis and treatment. Image quality assessment (IQA) methods are useful for image security and authentication by detecting whether an image has been tampered or altered. A No-Reference Image Quality Assessment (NR-IQA) method is proposed in this chapter for image security and authorization without the need of a reference image. The proposed methodology works on the hypothesis that distortion present in an image changes the naturalness of the image features. The unnaturalness of the medical images is found in wavelet domain by fitting the sub-band coefficients distribution using a Normal Inverse Gaussian (NIG) Probability Density Function (PDF). The noise feature is calculated by fitting a Gaussian PDF in diagonal sub-band. Further, the Natural Scene Statistics (NSS) and spatial correlations features are calculated. Support Vector Machine Regression (SVR) is implemented for classification and regression to estimate the quality factor. The experimental results show the performance improvements of the proposed method are 0.84%, 1.12%, and 5.04% over the state-of-the-art NR-IQA methods in terms of SROCC, PLCC, and RMSE parameters, respectively, on various distortions including JPEG 2000 (JP2K), JPEG, additive white noise (WN), Gaussian Blur (GB), and Fast Fading (FF).

Pages: 19

Since the advent of digital cameras, many image editing tools have spread and allowed everyday users to modify picture content. Image filtering and retouching are broadly used to enhance color, contrast, and brightness to improve pictures’ aesthetics. Going deeper into image editing, Copy Move is considered one of the most popular alterations in the pixel domain. It consists of copying and pasting a portion of pixels onto another area in the same digital image. Having counterfeited pictures travel all over the Internet is pretty usual nowadays. However, spotting Copy Move is not straightforward to human eyes, as editing tools have achieved such high accuracy and detail. That is often used for two reasons: fun or criminal purposes. The former represents a way to hide or add multiple copies of an object, a face, or whatever lets users have fun. In that case, the latter undermines the trustworthiness of digital images as proof of something taken from a visual scene. This chapter surveys this field’s SOTA (state-of-the-art) methods by accounting for image features, approaches, and datasets used for experimental campaigns. In particular, low-level image processing, machine learning, and deep learning methods have been analyzed and compared against some publicly available datasets, including different tampering transformations, such as translation, rotation, and scaling. Finally, some considerations are discussed on the pros and cons of SOTA methods.

Pages: 28

By managing Internet of Things devices, smart home automation solutions enable consumers to transform their homes into smart homes. Smartphone apps enable users to manage and centralize smart devices. Voice convenience makes sharing private information easy to home Internet of Things speech systems like Amazon Alex. It is difficult to search for voice while protecting user data. However, Alexa, Siri, and Google Assistant are gaining popularity. Today, data security is a major concern. This chapter describes a Multi-Layer Encryption with Advanced Encryption Standard scheme that employs ME-AES and Twofish encryption algorithms. ME-AES is proposed as a multi-layer encryption method for intelligent assistants. Intelligent assistant requests are scrambled and encrypted using Identity-Based Encryption (IBE) and BLAKE2 hashing keys. User email addresses and device serial numbers are accepted by the IBE software. Normal data is encrypted with AES, while sensitive data is encrypted with AES plus Twofish. This multilayer architecture safeguards data in a timely and efficient manner However, our proposed system, ME-AES, outperforms the other existing methods with an overall accuracy of 98.45%, which is higher than the other methods. And its encryption time is 4.872 seconds, which is the shortest time taken when compared to other recent work.

Pages: 24

In today’s technologically driven era, fingerprint biometric systems are widely used systems to provide secure access control and to protect the sensitive data of users. In fingerprint biometric systems, minutiae points are mainly employed to compute the template that will be stored in the database for recognition purposes. However, the stored template is vulnerable to attack and can be used to reconstruct the fingerprint image as discussed in the literature, which eventually causes an individual’s permanent identity loss. Therefore, the fingerprint template should be protected before storing it in the database. To do so, cancelable biometrics-based approaches are one of the prominent approaches presented in the literature. The main principle of cancelable biometrics is to store the transformed secure template in place of the original template, which makes reconstruction of fingerprint images an infeasible task. Considering the importance of cancelable biometrics, this chapter presents a detailed discussion of the concept of cancelable biometrics and a brief discussion of various recent existing techniques. Furthermore, a detailed discussion of a few of the recent cancelable fingerprint template generation approaches is also presented in this chapter, which provides a comprehensive understanding of these approaches, along with a discussion on their experimental results.

Pages: 19

With the advent of Industry 4.0, there are now unprecedented possibilities and challenges to be pursued across numerous domains. However, data privacy preservation is one of the key challenges that needs to be addressed, especially in sensitive domains such as the medical field. This is a critical concern, as the privacy of personal data is of utmost importance, and encryption, watermarking, other data-hiding techniques, and differential privacy are different ways to address this. Most of these techniques need a secure and optimal key that is hard to break by adversaries. Computational intelligence (CI) techniques have rapidly gained popularity as a mainstream research area in artificial intelligence, robotics, and optimization. These intelligent techniques are increasingly used to solve complex optimization problems across various real-world applications. Therefore, this chapter showcases the secure key generation process as a complex optimization process. It also highlights the importance of data privacy, along with challenges, security measures, and the benefits of CI for data privacy preservation, followed by possible future directions.