摘要
利用小波变换方法进行图像噪声滤波是使用小波分析理论进行图像处理应用的一个重要内容.本文以MATLAB为平台,给出了最常用的4种基于小波变换的图像噪声滤波方法并进行了仿真效果比较,同时,针对模极大值滤波方法实现时的烦琐和速度慢问题,提出了中值滤波和模极大值滤波相结合的算法改进,简化了运算,提高了运算速度并保证了滤波质量.试验得出小波域尺度相关性和改进后的小波系数模极大值法效果最理想.同时对小波变换方法的进一步改进与完善具有实用意义.
Using wavelet to filtering noises on image is an important application of wavelet analysis in image processing. The principle of four classical wavelet methods are described in this paper, and also the comparing them with implemented on the platform of MATLAB. Put forward an improvement of modulus maximum method to make operation easier and faster and make sure the quality of filtered image. And experimental results show that the methods of modulus maximum and scale relativity are better at filtering noises. The results given here is very useful for improvement and consummate of the wavelet.
出处
《天津理工大学学报》
2007年第3期35-37,75,共4页
Journal of Tianjin University of Technology
基金
天津市科委重大攻关资助基金(013180611)
天津市高等学校科技发展基金(20010204)
