基于复数小波变换增强带噪图像的空间自适应方法

A Complex Wavelet Based Spatially Adaptive Method for Noised Image Enhancement

针对目前的多尺度增强方法一般很难实现抑制噪声和凸显细节间有效均衡的问题,提出一种基于复数小波变换增强图像方法,充分利用复数小波兼具平移不变性和方向选择性的优势,首先通过相邻两层小波系数的相关性来有效区分噪声和图像边缘,并根据各层小波系数的分布设置局部阈值抑制噪声;在此基础上,自适应地选取增强函数来增强较弱的细节并保护原图像中的清晰边缘不产生失真.实验结果表明,运用该算法增强带噪图像可以在较好地抑制噪声的同时,显著地放大细节特征.

A new image enhancement method aimed at optimizing contrast of image features while minimizing image noise is proposed in this paper based on complex wavelet transform. The denoising is accomplished by using the correlations between different stages of wavelet coefficients, and then the subtle features are retained while noise is efficiently suppressed. Image enhancement is realized by a scale adaptive strategy that can emphasize features with low contrast while protecting the strong contrast features from distortions by over enhancement. Results of experiments show that this method offers significantly improved performance over the conventional and previously reported multi-scale enhancement algorithms.