Method of Structural-Statistical Coding of Video Segments in Spectral-Cluster Space
Andrii Krasnorutsky, Vitalii Kolesnyk, Anatolii Berchanov, Vladimir Barannik, Nataliia Kharchenko, Oleksander Malko · 2022
The article provides a general description of the problems of ensuring the necessary quality indicators of remote video service provision. The reason for this is the imbalance between the reliability of the received video information and time delays in the process of delivering video information over wireless communication channels. The destruction of existing wireless infocommunication technologies, which is the cause of disruption of the decision-making process in critical infrastructure systems, is briefly disclosed. It is proposed to solve the imbalance of video information delivery by using a set of technological solutions. The main component here is the use of technologies to reduce the bit rate of the video stream. The disadvantages of existing methods of reducing the bit rate of the video stream are substantiated. Among such disadvantages of the methods is that they have low efficiency: in terms of bit rate reduction and resistance to channel interference. It is proposed to solve the imbalance of reducing the bit intensity of the video stream by coding video data as part of integrated image processing technologies. Such technologies provide: additional consideration of new types of redundancy that are not related to the exclusion of psycho-visual redundancy; the construction of code structures with greater resistance to the negative impact of channel errors on the process of decoding video data. The technology of structural-statistical coding in the spectral-cluster space of the image transform is proposed. This technology is based on the principle of the duality of transformant components. The clustering of image transforms is proposed to be carried out in the binary space of their components. This creates conditions for simplifying the process of evaluating information characteristics and the formation of code values in binary space. As such information characteristics, it is proposed to use the number of series of units in the binary representation of the transformant components. The stages of the technological process of structural-statistical coding in the spectral-cluster space of the image transform are revealed.