Abstract: In order to meet the real-time video coding in embedded system and mobile device, a scalable complexity-distortion optimal motion estimation algorithm is proposed. The proposed algorithm can realize the computational complexity scalability of video encoder by adjusting complexity control parameters, and it can get the best video quality and obtain the relative optimal balance between the coding efficiency and distortion. Experimental results show that the proposed method can adjust the computational complexity of the encoder automatically according to the change of the processor’s computation ability. It can reduce the calculation expenses of motion estimation greatly, with less performance degradation and bit-rate increase.

Key words: H.264, Complexity-distortion, Motion estimation, Scalability

摘要： 为了满足嵌入式、移动设备的实时视频编码应用，提出一种可分级的复杂度-失真最优的运动估计算法，通过调整复杂度控制参数来实现视频编码器计算复杂度的可分级性，确保视频质量达到最优，在编码效率最高和失真最小间取得相对最佳平衡。实验结果表明，随着处理器计算能力的变化，该算法能自动调整编码器的计算复杂度，大大减少运动估计的运算量，同时图像质量和码率变化不大。

关键词: H.264, 复杂度-失真, 运动估计, 可分级

CLC Number: 

• TP391