An adaptive selection algorithm of macroblock coding candidate modes

In the macroblock coding mode selection process, to improve the judging speed of intra macroblock refresh, the codec usually uses the previous frame temporal information and low-level SVC layer image information to prejudge to reduce the range of candidate macroblock coding modes. As the compressed bitstreams are particular sensitive to error, in order to enhance the error resilience of compressed bitstreams, you can select some inter macroblock mode as the intra macroblock refresh in order to enhance the error resilience performance, to prevent the temporal error propagation. This paper firstly optimizes macroblock level rate-distortion model, and then proposes a candidate modes selection method which compares the rate-distortion function value of intra macroblock mode and inter macroblock mode, to reduce the number of candidate patterns of the rate distortion optimization, to reduce the complexity of the algorithm in the macroblock level, meantime to improve the intra macroblock refreshment rate, effectively prevent t propagation. Under the same bit error rate conditions, the computational complexity is reduced by 15%, and decoded image quality is no significant degrade.