Abstract: This paper proposes a rate control MGS optimization model based on greedy algorithm, which works in low complexity. The importance of each MGS slice layer is expected, and different rate-distortion models are proposed according to the frame type and quality layer type. The optimization scheme for MGS slices based on greedy algorithm is proposed. Experimental results show that the method distinctly improves decoded image quality, its average PSNR increases 0.5 db.

Key words: MGS coding, slice layer, rate control, rate-distortion model, greedy algorithm

摘要： 提出一种基于贪婪算法的MGS可伸缩编码的码率控制优化方法。该方法以较低的计算复杂度优化MGS编码，根据不同帧类型和质量层类型的统计特性分别建立不同的率失真模型，并基于贪婪算法确定MGS的片层优化分配方案。实验结果表明，该方法在相同码率下明显提高解码图像的质量，平均峰值信噪比可提高0.5 dB。

关键词: MGS编码, 片层, 码率控制, 率失真模型, 贪婪算法

CLC Number: 

• TP391