一种基于二值化和亚采样的文本图像压缩方法  

Compressing method of texts via bi_value and sub-sampling

作　　者：段崇雯[1] 侯臣平[1] 

机构地区：[1]国防科技大学理学院,湖南长沙410073

出　　处：《计算机应用》2005年第1期93-95,共3页journal of Computer Applications

基　　金：国家自然科学基金资助项目(10171109);国防科技大学本科创新基地基金

摘　　要：文中研究了彩色文本图像的高倍压缩方法。首先利用灰度窗口变换和整体阈值将原图二值化;然后进行亚采样处理并编码;在编码过程中,以经典的哈夫曼及游程编码为基础,设计了一种特殊的码流传输方式;最后用不同的插值方法近似的恢复了图像。实验结果证明,利用该方法当压缩比达到100倍以上时,恢复图像仍然具有高度的保真性。Two kinds of pre-procession methods were briefly discussed. Firstly, the true-color pictures are changed into bi-value ones, with the black pixels standing for the texts. Secondly, sub-sampling is applied to raise the compression rate (CR) to as about twice or more times as before. For compressing, a method called Modified Huffman Coding, which is a combination of traditional Huffman and Run-length coding method, is applied. After receiving the specially designed code stream, the decoding work reconstructs the pictures with several interpolation algorithms. Experimental results show that the reconstructed pictures are visually perfect with the size no more than 1% as original ones.

关 键 词：MHC编码 预处理 采样 插值 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]