经典的乡农编码，算法比较简单，了解乡农编码的请顶。

视频编解码，H.264编解码在视频会议中应用-Video codec, H.264 codec in the video conferencing application

本代码为数据压缩的C语言实现，它只是个学习程序，希望能够给参考者帮助！-the code for data compression C language, it is just learning the procedure, hoping to be able to help the reference!

JPEG图像压缩工具专辑-JPEG image compression tools album

mp3的经典 压缩 和 解压缩 源码。-mp3 of the classic compression source code.

DPCM编码的simulink仿真实现，很好很强大-DPCM coding simulink simulation, very good very strong! ! !

经过DCT变换用 LZW进行数据压缩 再还原回来-After DCT transform data using LZW compression and then restored back

This file contains a new and improved version of the Huffman coder, (June 29. 2001). The name is Huff06.m. There are also some additional files which are helpful when using Matlab for data compression: quantizer, different variants of run-length-encoding and end-of-block coding in Mat2Vec, and a program which do JPEG-like entropy coding. A complete compression example is shown in TestMat2Vec.m. This file is all you need for Huffman coding in MatLab.-This file contains a new and improved version of on of the Huffman coder, (June 29. 2001). The name is Huff06.m. There are also some additional files which are helpful wh en using Matlab for data compression : quantizer. different variants of run-length-encoding an d end-of-block coding in Mat2Vec. and a program which do JPEG-like entropy coding . A complete compression example is shown in Tes tMat2Vec.m. This file is all you need for Huffma n coding in MatLab.

使用matlab开发的jpeg工具，其中没有使用matlab中包含的字母进行预测，量化都是自己写的，为lena做压缩，但最后没有做熵的过程

又一个关于压缩的东东，是源码哦，本来用来研究winzip的-compress gzip

In each step the LZSS algorithm sends either a character or a pair. Among these, perhaps character "e" appears more frequently than "x", and a pair of length 3 might be commoner than one of length 18, say. Thus, if we encode the more frequent in fewer bits and the less frequent in more bits, the total length of the encoded text will be diminished. This consideration suggests that we use Huffman or arithmetic coding, preferably of adaptive kind, along with LZSS.-In each step the LZSS algorithm sends either a character or a lt;position, lengthgt; pair. Among these, perhaps character "e" appears more frequently than "x", and a lt;position, lengthgt; pair of length 3 might be commoner than one of length 18, say. Thus, if we encode the more frequent in fewer bits and the less frequent in more bits, the total length of the encoded text will be diminished. This consideration suggests that we use Huffman or arithmetic coding, preferably of adap