The invention discloses a method for evaluating the quality of a
stereo image based on binocular
information processing. The method comprises the steps that simple
cell information matrixes are obtained through
complex wavelet transform and
complex cell information matrixes are formed through collection of coefficients of
complex wavelet transform; on the basis of the
complex cell information matrixes, two aspects are evaluated, in the first aspect, the binocular
complex cell information matrixes are blended, so that a binocular blended image is obtained, and according to the fact that the human eyes has different degrees of visual sensitiveness to binocular information in different frequencies, the evaluated quality of the binocular blended image is finally obtained; in the second aspect,
binocular stereo perceptual information is extracted from each complex
cell information matrix, with the consideration of the stereoscopic
parallax angle and with the combination of the binocular masking characteristic, the human eyes can be influenced by the background luminance, the texture
mask, the space
contrast ratio and the like in the
visual perception process, and the evaluated quality of stereo vision
perception is obtained; finally, with the combination of the evaluations obtained from the two aspects, a final
stereo image quality
evaluation result is obtained and the uniformity of the
evaluation result and the subjective perceptive performance of the human eyes is good.