论文部分内容阅读
为降低DCT域分形图像压缩的编码复杂度 ,文中引入了平坦块的概念 ,对那些经过DCT变换后 ,能量非常集中的块不做块匹配 ,而是直接编码输出。同时经统计发现 ,对于给定的区块 ,与它匹配最好的域块通常位于这个区块的上方或附近 ,文中利用这个事实构造了一种优化的自适应搜索方案 ,大大减少了编码所需要的时间。实验表明 ,与一般的DCT域分形图像压缩方法相比 ,文中方案在提高了编码速度的同时降低了比特率。
In order to reduce the coding complexity of fractal image compression in DCT domain, the concept of flat block is introduced in this paper. After DCT, the block with very concentrated energy is not matched, but directly encoded and output. At the same time, it is found through statistics that for the given block, the best matching block is usually located above or near this block. This paper uses this fact to construct an optimized adaptive search scheme, which greatly reduces the number of codes time needed. Experiments show that, compared with the general DCT fractal image compression method, the proposed scheme reduces the bit rate while increasing the encoding speed.