基于改进EZW的感兴趣区域图像压缩算法仿真  

The Interested Region's Image Compression Algorithm Based on Improved EZW

作　　者：刘利章[1] 杨艺山[2] 史浩山[1] 

机构地区：[1]西北工业大学电子信息学院,陕西西安710072 [2]温州大学信息学院,浙江温州325000

出　　处：《计算机仿真》2006年第9期191-193,314,共4页Computer Simulation

基　　金：国防基础研究项目资助

摘　　要：针对公路收费站牌照拍照系统对于图像压缩的特殊要求,该文提出了一种静止图像的感兴趣区域压缩算法。该方法通过引入快速逐次逼近量化和对小波变换边缘子带的特殊处理,对EZW算法进行了改进。并以改进的EZW算法为基础,对感兴趣区域的图像进行高质量编码,而对背景图像进行粗糙编码。该方法对于那些对感兴趣区域的图像编码质量要求较高,而对背景图像则要求较低的应用场合非常有效。经实验证明,该方法既能够保证感兴趣区域的高质量图像压缩,又节省了大量的存储空间。For special need of image compression in the photograph system of the charge station of the road, this paper introduces an interested regiong compression algorithm for still image. The algorithm improves the EZW algorithm through inducting fast successive approximation quantization and effective treatment of marginal subband of wavelet transform. Based on improved EZW, the algorithm encodes the interested region of the image with a better quality, and does it with a low - grade quality in the unimportant region. The algorithm of the paper is effective to the applications that have better quality in the interested region of the image and low quality in the unimportant region of the image. Under the limited storage space condition, the image encoded by the algorithm not only meets the limit of the storage space, hut also ensures the reconstructed image with a better quality.

关 键 词：感兴趣区域 小波变换 子带 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]