Abstract

Nowadays, we need to compress and decompress many types of images such as natural, PC integrated, medical ones ... Depending on the characteristic of images, PSNR values can vary up to 30 dB with the same lossy coder. How to establish a new relationship between the features of different types of images and the lossy coding performance PSNR? In this paper, we introduce some measures of image characteristic which are called CI (characteristic of image) and build a mathematical function of the relationship between CI and PSNR.