摘要
为了更好地恢复图像，利用小波变换的思想，提出了一种变分和小波变换相结合的图像去噪算法。该算法的思想是先构造一个用带有韦伯心理学的范数估计图像正则性的变分泛函，然后在小波域中最小化变分泛函得到还原图像。与传统的直接求泛函最小化的问题有区别，该算法是用变分的思想再结合小波变换进行图像去噪。小波变换后的高频分量具有丰富的细节边缘信息，因而能够重构出高质量的图像，而且小波的引入使得新算法具有运行时间短、速度快的特点。理论分析和实验仿真表明，该算法能达到比单一方法更好的恢复效果。
To recover the image, it proposes an algorithm combining TV model and wavelet transform for image denoising. The main plan is to put forward a Weberized TV functional for estimating the regularity of the image, and minimize the TV functional in the wavelet domain in order to achieve the recovered image.The direct difference compared with the minimum of the traditional functional,the algorithm is re-thinking the variational wavelet transform for image denoising.High frequency components of wavelet transform have a wealth of detail edge information, so it can reconstruct high quality images, and the introduction of wavelet algorithm makes the text short and fast in the new running time.Theoretical analysis and simulation show that the algorithm can recover better than the effect of a single method.
出处
《计算机工程与应用》
CSCD
北大核心
2011年第28期21-23,共3页
Computer Engineering and Applications
基金
国家自然科学基金（the National Natural Science Foundation of China under Grant No.10771213）
关键词
韦伯变分泛函
小波变换
图像去噪
Weberized total variation
wavelet transform
image denoising
