Detection and concealment of transmission errors in MPEG-2 images-a genetic algorithm approach

• Authors: Han-Chiang Shyu;Jin-Jang Leou
• Venue: IEEE Transactions on Circuits and Systems for Video Technology
• Year: 1999

The detection and concealment of transmission errors in MPEG-2 images using genetic algorithms (GAs) is proposed. For entropy-coded MPEG-2 images, a transmission error in a codeword will not only affect the underlying codeword but also may affect subsequent codewords, resulting in a great degradation of the received images. Here, a transmission error may be a single-bit error or a burst error. The objective of the proposed approach is to recover high-quality MPEG-2 images from the corresponding corrupted MPEG-2 images without increasing the transmission bit rate. In the proposed error-detection approach, by using the constraints imposed on compressed image data, all the slices within an MPEG-2 picture can be correctly located. After a slice is located, similar to Chu and Leou (1998), transmission errors within the slice are detected by two successive procedures: 1) whether the slice is corrupted or not is determined by checking a set of error-detection conditions under decoding and 2) the precise location (block-based) of the first transmission error (i.e., the first corrupted block within the corrupted slice is located by a block-based backtracking procedure. For a corrupted block, the proposed GA approach to error concealment is employed to conceal the corrupted block by iteratively performing reproduction/crossover/mutation operations and evaluating the proposed fitness function until the stopping criterion is satisfied. Based on the simulation results obtained in this study, the proposed approach can recover high-quality MPEG-2 images from the corresponding corrupted images up to a bit error rate of 0.5%