National Natural Science Foundation of China (No.61071091);Youth Fund of National Natural Science Foundation of China (No.61001152);Graduate Research Innovation Program of univerities in Jiangsu Province (No.CX10B_188Z);Research Foundation of Nanjing University of Posts and Telecommunications (No.NY210069)
TANG Gui-jin, ZHU Xiu-chang, LIU Tian-liang. An Error Concealment Algorithm for Stereoscopic Images Based on Local Reliable Disparities[J]. Acta Electronica Sinica, 2013, 41(4): 781-786.
DOI:
TANG Gui-jin, ZHU Xiu-chang, LIU Tian-liang. An Error Concealment Algorithm for Stereoscopic Images Based on Local Reliable Disparities[J]. Acta Electronica Sinica, 2013, 41(4): 781-786. DOI: 10.3969/j.issn.0372-2112.2013.04.026.
An Error Concealment Algorithm for Stereoscopic Images Based on Local Reliable Disparities
In order to minimize the transmission errors for stereoscopic images
we propose an error concealment algorithm based on local reliable disparities.First
taking the block loss and the factors of color similarity and geometric distance proximity into account
we design a base-point-biased window to conduct the adaptive weight disparity matching.Second
local reliable disparities are calculated according to the disparity constancy and left-right consistency.Finally
the Winner-Takes-All (WTA) strategy is adopted to estimate the disparity of lost block which will be used to get the data of corresponding block for error concealment.Experimental results show that compared with other algorithms
the proposed method can produce better PSNR and subjective quality of reconstructed image while the computational complexity is almost the same.