Stereo image coder based on MRF analysis for disparity estimation and morphological encoding

Abstract

We present a stereoscopic image coder based on the MRF model and MAP estimation of the disparity field. The MRF model minimizes the noise of the disparity compensation because it takes into account the residual energy, smoothness constraints and the occlusion field. The disparity compensation is formulated as a MAP-MRF problem in the spatial domain and the MRF field consists of the disparity vector and occlusion field, which is partitioned into three regions by an initial double-threshold setting. The MAP search is conducted in a block-based sense on one or two of the three regions, providing faster execution. The reference and the residual images are decomposed by a discrete wavelet transform and the transform coefficients are encoded by employing the morphological representation of wavelet coefficients algorithm. As a result of the morphological encoding, the reference and residual images together with the disparity vector field are transmitted in partitions lowering the total entropy. The experimental evaluation on synthetic and real images shows beneficial performance of the proposed algorithm.