Description: 熵编码(entropy encoding)是一类利用数据的统计信息进行压缩的无语义数据流之无损编码。本章先介绍熵的基本概念，然后介绍香农-范诺(Shannon-Fano)编码、哈夫曼(Huffman)编码、算术编码(arithmetic coding)、行程编码(RLE)和LZW编码等常用的熵编码方法。 哈夫曼编码建议了一种将位元进位成整数的算法，但这个算法在特定情况下无法达到最佳结果。为此有人加以改进，提供最佳整数位元数。这个算法使用二叉树来设立一个编码。这个二叉树的终端节点代表被编码的字母，根节点代表使用的位元。除这个对每个要编码的数据产生一个特别的表格的方法外还有使用固定的编码表的方法。比如加入要编码的数据中符号出现的机率符合一定的规则的话就可以使用特别的变长编码表。这样的编码表具有一定的系数来使得它适应实际的字母出现机率。-Entropy coding (entropy encoding) is a kind of use data of statistical information compression without semantic data flow of nondestructive coding. This chapter first introduces the basic concept of entropy, and then introduce Shannon- occupies (Shannon- Fano) coding, Huffman (Huffman) coding, arithmetic coding (arithmetic coding), stroke encoding (RLE) and LZW encoding used entropy coding method. Huffman encoding suggest a bit carry into integer algorithm, but this algorithm in specific cases not to achieve the best results. Therefore some improved, providing the best integer bit number. This algorithm using binary tree to set up a code. The binary tree terminal node representing the code letters, on behalf of the root node used bits. In addition to this for each to coded data to create a special form of the method and use fixed code table method. Such as join to coding data symbols appear probability accord with certain rules words can use special variable ChangBian chart. This code Platform: |
Size: 18688000 |
Author:杨飞帆 |
Hits:
Search - LZW data compression algorithm with source code