Search

In recent years, BCT has garnered significant attention from researchers worldwide. The technology in question is a distributed database system characterised by its decentralised nature and lack of reliability. BCT has been widely adopted by numerous governments and scholars across various sectors for a number of years. Blockchain technology also involves highly innovative and advanced concepts. Given the increasing interest among scholars in the academic community regarding the agrifood supply chain, the objective of this study was to investigate BCT and its potential for application in the fields of food and agriculture. This research paper presents a bibliometric analysis of articles on the utilisation of BCT in the fields of food and agriculture. This study discusses scholarly articles that have been published in esteemed academic journals and conferences. Through our bibliometric analysis, we aimed to discern the recurring trends and themes within the research on BCT in relation to agrifood systems. Furthermore, this study examines a diverse array of research domains, prominent scholarly publications, leading publishing platforms, prominent funding institutions, and the prospective trajectory of future research. This study also presents the prominent patterns and themes within this field through an analysis of the most influential scholarly articles, authors, countries, and keywords found in the existing literature. Hence, this research employed various analytical techniques, including analyzing the co-occurrence of author keywords, bibliographic coupling analysis, network view map analysis, and co-citation analysis. This study holds promise as a valuable learning resource for aspiring researchers seeking to acquire compelling and pertinent information about research outcomes from studies on the utilisation of BCT in the field of smart agriculture.

Multimedia security has become challenging due to large amount of content generation and its distribution over network. Copyright protection and content authentication are the major key factor that avoids illegal distribution of digital data. However, due to availability of high bandwidth network, copyright violation is very common and many copies of data can be illegally distributed over network. Thus, to ensure multimedia security, image watermarking methods have been introduced which is a kind of information hiding technique. Watermarking provides an effective way to ensure copyright protection and content authentication and can be implemented in spatial and transform domain. The use of wavelet transforms in watermarking increases the embedding capacity and enhances the imperceptibility of the watermarked image. Being motivated from the use of wavelet transforms in image watermarking, in this chapter, a hybrid combination of the wavelet transforms have been discussed for single and multiple image watermarking. The transforms, combined in this chapter, are nonsubsampled contourlet transform (NSCT), discrete cosine transform (DCT) and multiresolution singular value decomposition (MSVD). This hybrid combination will take advantage of the characteristics and NSCT, DCT and MSVD to build a robust watermarking system in wavelet domain against signal processing and geometrical attacks.

Fusion of multisensor images has shown a potential application in various application domains such as security, medical imaging etc. The recent developments in medical imaging sensors have been a great motivation for fusion due to their complementary nature. This chapter aims to address medical image fusion in complex wavelet domain and provides a detailed study of fusion methods. The wavelet transforms based fusion methods are ahead of other methods in terms of signal representation, complementary information and redundancy. These properties make wavelet transforms suitable for multisensory image fusion. The fusion experiments have been demonstrated over several sets of medical images for different fusion rules in complex wavelet domain. Visual and quantitative evaluation of the proposed fusion results with state-of-the-art fusion methods showed the effectiveness and goodness of the complex wavelet transform based fusion methods.

The large availability and access of video data for education, entertainment such as games and movies poses several challenges of copyright violation and illegal distribution of data. More specifically, pirated video distribution has been a key threat for such copyright violation. To handle these challenges, watermarking techniques can be used to provide copyright protection. The wide application of wavelet transform in the implementation of image watermarking techniques compelled to discuss wavelet domain video watermarking techniques in this chapter. The present chapter provides an overview of video watermarking techniques in wavelet domain and discusses various methods used for embedding and extraction of watermark. It also provides the description of various attacks and key characteristics of video watermarking techniques with comparative analysis of results for evaluation of the existing methods.

This chapter proposes a new method of camouflaged person identification that combines discrete wavelet coefficients with support vector machine (SVM) classifier. Multiresolution property of wavelet transform provides invariant person identification against camouflaged scenes and do not get affected by similar foreground and background objects. Flexibility of wavelet and SVM makes the proposed method robust while providing better efficiency. For evaluation of the proposed method, we have experimented it over CAMO_UOW dataset. From objective evaluation it is clear that proposed approach outperforms existing camouflaged person identification approaches. The proposed methodology is simple and does not depend on a specific special resolution. It is suitable for detecting camouflaged persons in the images or videos where foreground and background are almost similar.

The huge amount of multimedia data available over Internet compelled to adopt new methodologies day by day. It becomes extremely complex to process and filter data as per user requirements. Hence, new tools, technologies are being used one after another. In this era of artificial intelligence, where machine and deep learning has been evolved, there is a great requirement of intelligent multimedia processing to fulfill and meet the requirements of users. With this aim, this chapter discusses a few key challenges in the visual information processing that are directly related to the societal benefits. Its major components are healthcare, education, transportation and security. A methodological enhancement of the techniques are being discussed to provide future research directions to the multimedia applications in wavelet domain.

Wavelet transforms have become an important mathematical tool that has been widely explored for visual information processing. The wide range of wavelet transforms and their multiresolution analysis facilitate to solve complex problems ranging from simple to complex image and vision based problems. The present chapter aims to provide an overview of existing wavelet transforms ranging from classical to new generation wavelets. This chapter discusses the basics of the discrete wavelet transform (DWT) followed by new generation wavelet transforms and highlights their useful characteristics. Other than DWT, the present chapter provides a brief review on dual tree complex wavelet transform (DTCWT), curvelet transform (CVT), contourlet transform (CT), contourlet transform (CNT), nonsubsampled contourlet transform (NSCT) to provide fundamentals and understanding of the wavelet transforms.

This study demonstrates analysis of an advanced technique that enhances performance of facial expression recognition method. Face preprocessing is conducted by face cropping. These cropped faces are rescaled by using bilinear interpolation. Multi-scale wavelet is used for extraction of facial patterns. Extracted features are down-sampled using principal component analysis (PCA) to reduce execution time as well as misclassification. The approach reduced the image dimensions and preserved the perceptual quality of the original images. Downsampled features are classified using multiclass support vector machine (SVM) that has a one versus all architecture. This system is trained with benchmark JAFFE and CK+ facial expression datasets. Performance of wavelet is compared with other existing frequency domain techniques and wavelets are found better in terms of recognition rate.

This chapter analyses real world human activity recognition problem. It uses discrete wavelet transform and multiclass support vector machine (SVM) classifier for recognition. The experiments are done using Weizmann and KTH action datasets. Objective evaluation is done for nine activities walk, run, bend, gallop sideways, jumping jack, one handwave, two handwave, jump in place and skip from Weizmann dataset. Six activities that are considered from KTH dataset are handwaving, running, walking, boxing, jogging and handclapping. Results are shown qualitatively as well as quantitatively on two publicly available dataset Weizmann and KTH. Quantitative evaluations demonstrate better performance of the proposed method in comparison to the existing methods.

Facial expression recognition plays a significant role in human behavior detection. In this study, we present an efficient and fast facial expression recognition system. We introduce a new feature called W_HOG where W indicates discrete wavelet transform and HOG indicates histogram of oriented gradients feature. The proposed framework comprises of four stages: (i) Face processing, (ii) Domain transformation, (iii) Feature extraction and (iv) Expression recognition. Face processing is composed of face detection, cropping and normalization steps. In domain transformation, spatial domain features are transformed into the frequency domain by applying discrete wavelet transform (DWT). Feature extraction is performed by retrieving Histogram of Oriented Gradients (HOG) feature in DWT domain which is termed as W_HOG feature. For expression recognition, W_HOG feature is supplied to a well-designed tree based multiclass support vector machine (SVM) classifier with one-versus-all architecture. The proposed system is trained and tested with benchmark CK+, JAFFE and Yale facial expression datasets. Experimental results of the proposed method are effective towards facial expression recognition and outperforms existing methods.