Image error concealment using optimized local pixel matching and directional interpolation

In this paper, we present a new spatial error concealment algorithm with optimized local pixel matching and directional interpolation. First, we apply a variant version of dynamic programming to image pixels around a corrupted image region. This process provides best-aligned pixel pairs from boundary pixels around the corrupted image region to dynamically estimate local image textures. Then, with derived corresponding pixel pairs, we use directional interpolation to recover image pixels in the corrupted image region. With both optimized local matching algorithm and directional interpolation, we match pixels using the pixel sequences from the upper and lower of the lost region, the pixel sequences from the left and the right of the lost region, the pixel sequences from the upper and the left of the lost region, and pixel sequences from the lower and the right of the lost region. We select the best matching based on the set up criterion and use them for error concealment. From experimental results, the proposed spatial error concealment provides the enhancement of the dynamic capbility in capturing the image texture comparing to previous works as well as subjective and objective quality improvement of concealment results in recovering the lost region of images comparing to previous works.