基于边缘方向改进的自适应窗口Lee滤波算法

摘要

为克服自适应窗口Lee滤波算法对图像边缘方向不敏感的弊端,提出了一种改进方法.其基本思路是先对图像纹理区进行分类,然后在方向上使用改进的四向模板获取方向窗口,分别对不同的方向使用不同的窗口增长方式对图像进行滤波.实验表明,改进后的算法相比传统算法在有效去除噪点的同时保存了大量的边缘与细节信息.%In order to overcome the disadvantage of adaptive window Lee filtering algorithm insensitive to the image's edge orientation,an improved method is proposed. The basic idea is to classify the image te-xture region, and then use the four direction template to obtain the direction of window, finally filtering the image by different window growth modes in different directions. Experimental results show that compared with improved algorithm of traditional algorithm is effective in removing noise while preserving edge and detail information of a large number.