摘要
利用小波变换方法对图像进行去噪处理是小波理论在图像处理中的一个重要应用。针对传统的去噪算法对信号和噪声在不同的分解层的传播特性缺乏有效分析这一不足,文中介绍了提升方法的基本原理,给出了用提升原理构造传统小波的实现方法,并将Daubechies(9/7)提升格式小波应用到二维图像去噪过程中,提出了一种对小波分解后各个层次上的水平、垂直、对角方向上的高频系数矩阵进行分块处理的自适应阈值去噪方法。通过与传统算法对比,文中方法计算简单,运算速度快,去噪效果较好。
Using wavelet transform to filter noises on image is an important application of wavelet analysis in image processing. The uadi- tional denoising algorithms are lack of effective analysis in different layers for signal and noise＇ s propagation characteristics. In response to this shortage, a new threshold construction program is proposed based on lifting algorithm. Local contrast, level of wavelet decompo- sition and statistical properties of high-frequency coefficients are taken into account. The method makes block processing at high-fre- quency coefficient matrix for all levels of horizontal, vertical, diagonal direction. Experiments show that the method is simple, fast spoed, better denoising,
出处
《计算机技术与发展》
2012年第7期78-80,84,共4页
Computer Technology and Development
基金
国家自然科学基金(61075007)
西安理工大学基金(108-210901)
关键词
图像去噪
小波变换
提升小波
自适应阈值
image deniosing
wavelet transform
lifting wavelet
adaptive threshold
