Method of recurrent truncated-positional coding video segments in uneven diagonal space

Abstract

The subject of research in this article is methods of encoding transformed video segments to reduce their bit volume without loss of information integrity.  The goal is to develop a technology for coding uneven diagonal sequences under the conditions of their arbitrary positioning in the transformant. Task: to justify the approach of creating new methods of encoding video segments, considering the features of the combinatorial configuration of transformants; to create a method of formatting the coordinate system of spectral components in an uneven diagonal direction; to develop a method of encoding non-uniform diagonal sequences in a two-dimensional spectral space; to build a technology for recurrent realization of the process of sliding truncated-positional coding of uneven-diagonal sequences. The methods used are: mathematical models for estimating the number of structural-combinatorial and psychovisual-combinatorial redundancies in an uneven-diagonal spectral space; methods of positional coding. The following results were obtained. Potential advantages of considering the combinatorial configuration of the transformant based on its reformatting according to the non-uniform diagonal structure are substantiated. The technology of recurrent truncated positional coding of video segments in non-uniform diagonal space has been developed. It is based on two technological components. These include: the first component - a pyramidal system of positioning diagonals and their component in the transformant is created. The second component - a method of recurrent implementation of truncated-positional coding of uneven-diagonal sequences, is constructed. Such coding is organized regardless of the positioning of the diagonals in the two-dimensional spectral space of the transformant. Comparative evaluation revealed the advantages of the created method over standardized transformant coding methods. The advantage is achieved by the level of bit volume reduction and reaches an average of 15-30 %. Conclusions. For the first time, a method for establishing the coordinates of the components in the diagonals was developed. It is based on considering the features of the structural configuration of the transformant. This creates conditions for reducing time delays for processing video segments. For the first time, a method for recurrent coding of diagonals based on truncated-positional systems was created. This makes it possible to avoid the cases of the violation of the conditions of mutually unambiguous code conversion.