一种最优化链码指纹二值细化图像压缩编码  

Optimized chain code compression algorithm for fingerprint binary image

作　　者：李超[1] 杜赓[1] 杨义先[1] 钮心忻[1] 

机构地区：[1]北京邮电大学信息安全中心,北京100876

出　　处：《计算机应用》2006年第10期2357-2359,共3页journal of Computer Applications

基　　金：北京市自然基金资助项目(4042022;4062025)

摘　　要：提出了一种适合于对线状结构的条形纹线二值图像进行压缩的最优化Freeman链码压缩算法——Freeman差分链码Huffman编码。与传统的Freeman链码相比,提出的压缩算法是基于Freeman链码、差分编码和Huffman编码的一种混和编码方式。通过理论分析和在指纹二值细化图上的实验结果证明,对于指纹二值细化图像,本算法优于现有的链码压缩二值图像的算法,针对于线状结构的条形纹线二值图像,本算法也优于其他压缩算法。其平均码长为1.7651bits,低于8方向Freeman链码或者Freeman差分链码的3bits的平均码长。An optimized compression algorithm for linear structure stripe image was discussed in this paper, which is Freeman differential chain code Huffman coding. Compared to the traditional Freeman chain code, the proposed one is a hybrid encoding method based on Freeman chain code, differential code and Huffman code. Theoretic analysis and results from experiments on fingerprint binary images show that this algorithm is superior to other binary image compression algorithms, especially to fingerprint binary image compression. The average code length of the proposed one in this paper is 1. 7651bits, which is shorter than that of 8 orientations Freeman chain code or Freeman differential chain code whose average code length is 3 bits.

关 键 词：FREEMAN链码 二值图像 压缩编码 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]