基于边缘检测的提升小波图像去噪

Lifting Wavelet Image De-noising Based on Edge Detection

为了图像去噪的同时能很好地保持图像的细节,提出了一种基于边缘检测的提升小波图像去噪方法,先用Canny边缘检测方法得出原图像的边缘,提升小波去噪时就可用阈值去噪而不必担心损害图像的重要特征。其去噪步骤是:对噪声污染图像进行提升小波去噪;对含噪图像提取边缘图像;对去噪图像高频系数进行阈值处理进行重构变换;图像合并。实验结果表明,该方法不但可以保持图像的重要信息,而且能够提高去噪后图像的信噪比。

In order to preserve image details while canceling image noise, this paper presents a lifting wavelet image de-noising method based on edge detection, which adopts the method of Canny edge detection to get the edge of the image, and uses a threshold in lifting wavelet de-noising without damaging image＇ s characters. The steps are as follows ： Carry out lifting wavelet de-nosing on noise pollution image ; Extract edge image from noise image; Reconstruct high frequency coefficient of de-noising image by thresholding operation ; Combine images. Experimental results show that the method can not only keep important information of images, but also can improve signal-to-noise ratio.