Оптимизация арифметического кодера для сжатия изображений, полученных при дистанционном зондировании водных объектов

Abstract

Introduction. The fast program algorithm of arithmetic coding proposed in the paper is for the compression of digital images. It is shown how the complexity of the arithmetic coder algorithm depends on the complexity measures (the input size is not considered). In the course of work, the most computationally complex parts of the arithmetic coder algorithm are determined. Performance optimization of their software implementation is carried out. Codecs with the new algorithm compress photo and video records obtained through the remote probing of water bodies without frame-to-frame difference. Materials and Methods. In the presented paper, a selection of satellite images of the Azov Sea area was used. At this, the software algorithm of the arithmetic coder was optimized; a theoretical study was conducted; and a computational experiment was performed. Research Results. The performance of the software implementation of the arithmetic coder is increased by the example of the VP9 video codec. Numerous launches of reference and modified codecs were made to measure the runtime. Comparison of the average time of their execution showed that the modified codec performance is 5.21% higher. The overall performance improvement for arithmetic decoding was 7.33%. Discussion and Conclusions. Increase in the speed of the latest digital photo and video image compression algorithms allows them to be used on mobile computing platforms, also as part of the onboard electronics of unmanned aerial vehicles. The theoretical results of this work extend tools of the average-case complexity analysis of the algorithm. They can be used in case where the number of algorithm steps depends not only on the input size, but also on non-measurable criteria (for example, on the common RAM access scheme from parallel processors).