摘要
基于小波变换的图像压缩算法在含噪图像和较低码率时出现的边缘模糊现象多年来一直未能得到很好的解决,为了解决这一问题,提出一种基于边缘检测的噪声图像压缩编码方法。首先,对图像进行小波边缘检测,确定哪些小波系数是图像的边缘特征,将其保护起来;然后,对小波变换域系数采用软阈值收缩方法实现去噪;最后,利用等级树集合分裂算法(SPIHT)算法对图像进行压缩编码。实验结果表明,本文提出的方法不仅能获得较高的图像压缩率、较好地去除噪声,而且能在一定程度上解决边缘模糊问题。
Edge blurring phenomenon, which occurs in the wavelet-based image compression algorithms under noised image or low bit rates, remains an open question for many years. Therefore, this paper presented a method of noise image compression and coOing based on edge detection. Firstly, the image was wavelet edge detected, the detected wavelet coefficients were protected, and then the noise was removed by soft-threshold method, and at last we compressed and coded it by using Set Partitioning in Hierarchical Trees （SPIHT） method. The experimental result shows that this method not only achieve high compression ratio, remove noise well but also solve edge blurring problem better.
出处
《计算机应用》
CSCD
北大核心
2008年第9期2297-2299,2305,共4页
journal of Computer Applications
基金
吉林省科技发展项目(20030307)
关键词
图像处理
边缘检测
小波阈值去噪
压缩编码
识别与压缩
image processing
edge detection
wavelet threshold denoising
compressed coding
recognition andcompression
