摘要
本文首先介绍无失真图像编码方法,然后提出一种基于块方向的自适应无失真图像编码方法.该方法首先将图像划分为若干块,自适应地为每一块选择一个预测误差最小的方向预测器.其次,一种有效的霍夫曼编码技术被用来编码误差图像,它覆盖了较大范围的信息熵.最后,给出了实验结果及与其它无失真编码算法的比较.
In this paper, lossless image coding is first introduced, and then an adaptive lossless image coding method based on block directions is presented. In this method, an image is first partition into blocks and then for each block a direction predictor that results in minimum prediction error is select adaptively. After that, an efficient adaptive Huffman coding technique for coding residual image is given which covers a wide entropy range. Finally, the experimental results and comparisons to other lossless coding algorithms are tabulated.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
1998年第1期89-95,共7页
Pattern Recognition and Artificial Intelligence