A Tool for Joint Error-Resilience and Error-Concealment in Block-Based Video Encoders

This paper analyzes a joint error resilience and error concealment method (J-ER-EC) for video sequences. The method is based on a spatial-domain error concealment approach plus some redundant information, providing an edge-based sketch-description of the whole frame. The technique has its origin in well-known ideas concerning the use of side information to help video concealment: in J-ER-EC the redundant information is primarily used to recover corrupted intra-coded (I) frames but, thanks to error propagation reduction, the benefits are not limited to I-frames. J-ER-EC is tested with a H.264 codec in a simple scenario involving "deterministic" losses: the results allow to evaluate the percentage amount of redundancy inserted as well as the performance of the method with respect to existing resilience and concealment solutions, both in terms of visual quality improvement and of achieved PSNR