Avs Video Standard Decoder Side Error Concealment Technology

With the extensive use of multimedia communication technology, the video communication has become an indispensable part of our daily life and people have taken the quality of video service into account. However, in the actual process of video signal transmission, the loss and damage of data packet are inevitable due to the physical defects of the channel, which lead to a serious decline in the quality of the video stream. The error resilient and error concealment technology is an effective way to provide protection and recovery for the packets and ensure the quality of video image transmission.As a standard with our own independent intellectual property rights, AVS video coding standard almost has the same coding efficiency and video quality as the current international video coding standard H.264. In addition, AVS video coding standard is free from many kinds of patent fees and has simple patent licensing model. And for its late start, AVS only uses a simple error control strategy. So, the study of the error resilient and error concealment based on AVS coding standard is of great significance. Combined with the AVS official software reference model, this dissertation is focused on the error concealment method on the decoding side based on the AVS video coding standard and has done some researches on the algorithm improvement, comparison and simulation. In particular, it contains the following three aspects:1. Spatial error concealment algorithm. Based on the existing spatial error concealment technologies, such as bilinear interpolation algorithm and directional interpolation algorithm, this dissertation has proposed a self-adaptive spatial error concealment algorithm. By comparing and evaluating these algorithms in AVS software reference model, the self-adaptive algorithm has a better result.2. Temporal error concealment algorithm. This dissertation has chosen three existing algorithms for comparing and analyzing, such as zero motion vector algorithm, side matching algorithm and boundary matching algorithm. Considering the advantages and disadvantages of these existing algorithms, this dissertation has proposed an improved algorithm. The improved algorithm has taken the multi-weighted boundary matching technology and direction diamond search pattern for more precise search. From the recovery result, the improved algorithm has achieved our prospective goal.3. Based on the spatial and temporal algorithms, this dissertation has proposed an error concealment algorithm for AVS video coding standard. According to the frame type, this algorithm takes different algorithm between the spatial method and temporal method. By evaluating and comparing with the algorithm from H.264 software reference model JM, our algorithm has further improved the quality of the video sequences.