IMPROVEMENT TO MOTION ESTIMATION FOR HIGH-EFFICIENCY VIDEO CODING

Abstract

The paper deals with video coding based on the use of dependent sources, which is caused by the need to accelerate the coding process. The coding time is directly dependent on the computing power, which with the advent of new standards is not enough. Therefore, the article suggests the results of the study, which are designed to improve existing methods, as well as to continue the progress of scientific research in this area. The results consist of using an approach to compression based on coding methods of dependent sources, which can significantly reduce the complexity of processing on the side of the encoder. The results of the study were obtained using two coding algorithms for fixed and dynamic plots. For comparison, JPEG and, respectively, MPEG encoding standards were defined, as well as an improved algorithm that was considered by the authors in previous papers. It solves the problem of efficiently recovering intermediate frames on the decoder side. When evaluating the movement, search styles have a significant impact on the performance of the algorithm, both the search speed and the quality of the search. The search style should be designed considering the characteristics of the distribution of the motion vector probability (MVP) of the sequences of the real plot. The purpose of this study is to improve the coding efficiency of video sequences using the Diamond Search (DS) algorithm. The results show that DS is efficient and reliable, it can always give a higher search speed in different sequences than other algorithms.