Open AccessCOMPRESSION OF HIGH DYNAMIC RANGE OF SAR IMAGES
Author(s) -
Ю. И. Голуб
Publication year - 2018
Publication title -
sistemnyj analiz i prikladnaâ informatika
Language(s) - English
Resource type - Journals
eISSN - 2414-0481
pISSN - 2309-4923
DOI - 10.21122/2309-4923-2018-1-51-57
Subject(s) - computer science , image compression , compression (physics) , transformation (genetics) , range (aeronautics) , artificial intelligence , dynamic range , computer vision , dynamic range compression , data compression , image quality , measure (data warehouse) , high dynamic range , image (mathematics) , tone mapping , selection (genetic algorithm) , image processing , data mining , biochemistry , chemistry , materials science , composite material , gene , telecommunications
The paper presents results of our experiments on compression of the high dynamic range SAR images. The range is equal to 16-bit. Objectives of study were comparison of known approaches to compression of the high dynamic range images; selection of optimal parameters for compression algorithms, and selection of a no-reference measure for image quality assessment after compression. Tone-mapping transformations like gamma correction, Ashikhmin-operator, mu-transformation, as well as no-reference image quality assessment measures were tested. The results of the experiments are presented in the article. It was concluded that further research and analysis of various functions and approaches to compression of dynamic range of SAR images is necessary, since including in the article approaches do not give stable and positive results on all SAR images. It was also concluded that after transformation 16-bit images, it is very difficult to estimate which image is better, and it is necessary to use no-reference image quality assessment measure.