COMPRESSION OF NOISY GRAYSCALE IMAGES WITH COMPRESSION RATIO ANALYSIS

• Published in: Сучасні інформаційні системи Vol. 9; no. 2; pp. 68 - 74
• Main Authors: Kryvenko, Sergii, Lukin, Vladimir, Bondžulić, Boban, Stojanović, Nenad
• Format: Journal Article
• Language: English
• Published: National Technical University "Kharkiv Polytechnic Institute" 30.04.2025
• Subjects: better portable graphics; image complexity; lossy compression; noise intensity
• ISSN: 2522-9052
• DOI: 10.20998/2522-9052.2025.2.09

The object of the study is the process of compressing noisy images in a lossy manner by better portable graphics (BPG) encoder. The subject of the study is the method for adaptive selection of the coder parameter Q depending on noise intensity and image complexity. The goal of the study is to consider the basic characteristics of lossy compression of remote sensing images contaminated by additive white Gaussian noise with giving recommendations of preferable Q setting. Methods used: numerical simulation, verification for test images. Results obtained: 1) the dependencies of compression ratio on Q are monotonically increasing functions; 2) their characteristics are strongly dependent on noise intensity and image complexity; 3) dependencies of logarithm of CR on Q contain information on possible existence and position of optimal operation point for compressed noisy images; 4) compression ratios for large Q contain information on image complexity with low sensitivity to noise presence and intensity; 5) it is possible to get useful information from dependences of compression ratio on Q. Conclusions: the results of this research allow: 1) estimating image complexity; 2) adapting Q to noise intensity and image complexity.