An Adaptive Complexity Control Algorithm for AVS-M Transforming Quantization

To the condition of the complexity of AVS-M transforming quantization all-zero block detection algorithm changes with video content changes, according to the characteristics of adjacent video frame contents have dependence, a adaptive complexity control algorithm for AVSM transforming quantization is proposed. According to reasonable parameter choice, the algorithm can converge to the given computational complexity requirements no more than 25 frames at most. The range ability of computational complexity after converging does not exceed ±3%, SNR is reduced in 3.58% and the complexity of transforming quantization in encoding process remains virtually unchanged.