Downsample-based multiple description coding and post-processing of decoding

Multiple description coding is a new coding technique engineered to solve the problems with transmitting video stream on the unreliable networks. In this paper, the author implements the multiple description coding based on the spatial sampling approach, in doing so the mismatching problems commonly found in using multiple description coding is thus avoided. Also realized are two distinctive multiple description decomposition schemes suitable for video images with various texture characteristics. In addition, to improve the multiple description decoder performance, edge dynamic matching interpolation is adopted in decoding for post-processing purpose. The simulation results suggest that the proposed coding methods can be more efficiently applied in video transmission under various network circumstances, and the image SNR is evidently enhanced after the post-processing is completed in decoding process.