摘要
为了进一步提高数字图像无损压缩的效率,本文提出一种基于高效率视频编码(High Efficiency Video Coding,HEVC)框架的无损压缩算法。在该算法中,宏块中的像素不是作为一个整体来预测的,而是分成不同的部分先后进行预测。本文将像素分为3部分:第一部分的像素参考相邻宏块的像素来预测;其他部分的像素,利用前面已经预测的像素作为参考来进行预测,这样可以有效地去除空间冗余。另外,由于无损压缩中将HEVC框架中的变换量化去除,因此针对残差的分布特性,采用自适应扫描的方法将残差以合理的顺序进行熵编码,从而实现进一步提高压缩效率的目的。
In order to improve the efficiency of lossless compression, this paper proposes a lossless compression algorithm based on high efficiency video coding. In this algorithm, pixels in macro block are divided into different part to predict, instead of using a block-based prediction as a whole. More specifically, pixels are divided into three parts： pixels of first part are predicted by re- ferring the neighbor macro block pixe!s; then pixels in other parts are predicted using the pixels, which has been predicted, as reference, which can eliminate effectively spatial redundancy. Moreover, since the transform and quantization modules are re- moved from the framework of HEVC, the residuals need to be adjusted to the distribution characteristics in a reasonable order for the entropy coding, thus the purpose of further improving compression efficiency is achieved.
出处
《计算机与现代化》
2013年第8期23-26,共4页
Computer and Modernization
