Обробка відеопотоку

This article proposes three methods for parallelizing the compression of video streams from Earth surface scanners installed on microsatellites. The core innovation lies in the application of vertical, horizontal, and block-based segmentation of the input stream to enable scalable and high-throughput image compression. The study establishes a quantitative relationship between video stream intensity and satellite parameters, and also determines the required number of compression cores based on the satellite's characteristics and the performance of a single core.

The article explores the possibilities of implementing the JPEG-LS image compression algorithm on Field Programmable Gate Arrays (FPGA) for processing monochrome video streams from Earth surface scanners. A comparison of software implementations of the algorithms, their compression ratio, and execution time is conducted. Methods for improving FPGA performance are considered, using parallel data processing and optimized data structures to accelerate compression and decompression processes.