基于小波变换与改进的Roberts算子融合的图像边缘检测

Image edge detection based on wavelet transform and improved Roberts operator

本文提出了一种改进的Roberts算子,并在此算子边缘检测框架基础上,采用了一种小波多尺度边缘检测方法,对不同尺度下的小波变换子图像,通过改进的Roberts算子对子图像进行空间微分,得到对应尺度上的边缘图像,各尺度下边缘图像通过小波重构、融合运算得到最终的边缘图像。实验结果表明,该算法实现简单,能有效地抑制噪声,补偿弱边缘,可以有效获取多个尺度下的边缘信息,对边缘信息定位精度高,是一种可行、有效的图像边缘检测的方法。

An improved Roberts operator is proposed in this paper. Based on the framework of edge detection of the operator, a multiscale edge detection method with wavelet transform is adopted, the spatial first order differential are performed on the wavelet-transform sub-images with different scales through improved Roberts operator to get the edge images with corresponding scale, and the final edge image is derived from the wavelet reconstruction, fusion against the edge images in various scales .The experimental results show that the algorithm is simple to be performed and has superiorities in suppressing noise, compensating the weak edge and acquiring the edge information of different scales. And then this algorithm can position edge information exactly. It is a feasible and effective edge detection method.