Title :
Restoration of Highly Impulses Corrupted Color Image by Local Entropy Measure
Author_Institution :
Electr. Eng. Dept., Mingshin Univ. of Sci. & Technol., Shinchu, Taiwan
Abstract :
This paper proposes an effective scheme using local entropy measure to restore highly salt-pepper impulses up to 99% corrupted colour image. A local information measure to restore impulses corrupted image had been proposed before. Its restoration quality is not satisfactory at high impulse densities. The proposed algorithm presents significant quality of restored image than above proposed method not only for grey images but also for colour images. In addition, its preservation of edges and details is superior at high impulse densities, also, this scheme presents good empirical results by both the visual quality and quantitative performances in terms of peak signal to noise ratio (PSNR) and normalized colour difference (NCD).
Keywords :
entropy; image colour analysis; image restoration; NCD; PSNR; detail preservation; edge preservation; grey images; highly impulses corrupted color image; impulse corrupted image restoration; impulse density; local entropy measure; local information measure; normalized colour difference; peak signal to noise ratio; salt-pepper impulse restoration; visual quality; Entropy; Image color analysis; Image restoration; Nonlinear filters; PSNR; adaptive filtering window; boundary detection; entropy; local contrast; local information;
Conference_Titel :
Computer, Consumer and Control (IS3C), 2014 International Symposium on
Conference_Location :
Taichung
DOI :
10.1109/IS3C.2014.204