LANIC:一种同时支持无失真及高保真的图像压缩方法
LANIC: A LOSSLESS AND NEAR LOSSLESS IMAGE CODER
-
摘要: 提出了一种同时支持无失真及高保真的图像压缩方法—— L ANIC.该方法首先将图像划分为若干块 ,对于每一个图像块 ,根据边缘的强弱使用 DPCM编码器或者 DCT编码器 .DPCM编码器采用两级去相关技术和四叉树 Rice编码器 ;DCT编码器中 ,使用 DCT、Z形扫描、游程和变长编码 (VL C)技术 .实验结果表明 ,L ANIC的无失真压缩效果明显好于 JPEG无失真模式和 FEL ICS,略好于 CB9和 L OCO- I;高保真压缩模式下 ,当码率 =4~1bpp时 ,L ANIC的 PSNR均在 38d B以上 ,且高于 JPEG有失真模式Abstract: A lossless and near lossless image coder—LANIC is presented in this paper. In this method, the image is first partitioned into blocks. Then for each block, the edge detector automatically switches to DPCM coder or DCT coder. The edge blocks are coded by the DPCM coder that adopts a two level decorrelation scheme and a quartertree Rice coder. For non edge blocks, DCT, run length and variable length coding (VLC) are applied. The experimental results are as follows: Under lossless mode, LANIC significantly outperforms lossless JPEG and FELICS, and is a little better than CB9 and LOCO I; under near lossless mode, PSNR of LANIC between 4bpp and 1bpp is always above 38dB and better than the lossy JPEG.
下载: