Quality evaluation of impressed infrared images

As a kind of special images, infrared images have to be compressed for proper transfer and storage. The traditional method, which is based on MSE to objectively evaluate the quality of impressed images, sometimes cannot conform to effect of human vision perception. Edge mean square error is defined according to human vision system and a better evaluation result conforming is provided to human vision. Different thermal image compression algorithms are described. It is improved that edge mean square error is more reasonable to evaluate the images.