摘要
Sobel算子对图像进行边缘检测时,只对垂直与水平方向敏感,使得一些边缘检测不到,从而限制了Sobel算子的检测性能。针对此问题,文中提出了一种基于Gabor滤波器的边缘检测算法,先对图像进行Gabor变换,找出不同方向上人眼敏感程度低的系数,再用Sobel算子对变换过的图像进行边缘检测,最后把此图像与直接用Sobel算子检测的图像相加。实验表明,该算法改善了Sobel算子检测方向单一的不足,包含较多的边缘信息。
When using Sobel edge detection operator,it is only sensitive to vertical and horizontal direction and can' t make some edge detection,which limit the detection performance of the Sobel algorithm.According to this problem,this article presented an egde detection algorithm based on odd Gabor filter.Firstly,Gabor transform should be carried out on the image,to find out the different coefficient of human eye sensitivity,which is low,then detect the edge of the transformed image with Sobel algorithm.Finally,it and the image which detected by Sobel algorithm are added directly.The experimental results show that this algorithm modifies Sobel algorithm' s deficiency of singleness in edge detection,containing more testing information.
出处
《信息技术》
2014年第1期17-20,共4页
Information Technology
基金
河南省基础与前沿技术研究计划项目(102300410266)