基于统计假设测试的图像结构检测方法及修复

摘要

Image denoising is digital image processing one of the important methods, any image denoising algorithm will put some edge or texture image structure inevitably, based on the smooth off to repair the image denoising, firstly, according to statistics hypothesis testing principle given a measure image structure characteristics method, through image structure characteristics degrees can judge the local area is the edge area or pure noise texture, and then using the image area structure characteristic degrees to repair is to deal with the noise of image structure. Experiments show that this method has obvious effect, particularly for denoising excessive images have better effect.%图像去噪是数字图像处理中的一个重要方法，任何图像去噪算法都不可避免地会把一些边缘或纹理等图像结构给平滑掉，基于修复去噪后的图像中一些图像结构的目的，首先根据统计假设测试原理给出一种度量图像结构特征的方法．通过图像结构特征度可以判断出局部区域是边缘纹理区域还是纯噪声区域，然后应用图像结构特征度来修复被去噪处理后的图像结构。实验表明本方法具有明显的效果，尤其对于去噪过度的图像有更佳的效果。