Image evaluation algorithm for right view of stereoscopic video

Image quality assessment is an important problem image processing. In this paper, an image evaluation algorithm based on fuzzy theory is proposed to predict the quality level of right view image in compressed stereoscopic video. Block-based disparity estimation between stereoscopic videos is used in evaluation processing. With the help of inter-view correlation, the reconstructed left view at decoder replaces original right view as reference image to evaluate the quality level of the reconstructed right view image. Experimental results show that the proposed approach can predict the quality level of the right view image in stereoscopic video and the assessment result is well consistent with the perceptual characteristics of human eyes.