基于小波变换的图像边缘检测算法及像素结构分析

Structure Analysis and Improved Method of Images Edgedetection Based on Wavelets Transform

对基于多尺度小波变换的图像边缘检测算法进行了改进。通过处理小波变换后出现的负值解决了双边缘问题;通过处理两层邻域的像素点解决了边缘线间断问题,并与传统算法的结果进行了比较,结果表明,利用这一新的算法可使图像边缘更加光滑、边缘定位更精确。

A method of image edge detection based on wavelets multi-scale transform is presented through analysing sorts of image structure cells. By using negative value and dual point after wavelets transform can solve effectively image edge disconnection and rough. Compared with traditional method, it is shown that image edge is smoother and it is more accurate to make a reservation by using the method.