小波变换和Canny算子相结合的图像边缘检测算法

提出一种基于小波变换和Canny算子相结合的图像边缘检测算法.由于传统Canny算子在进行边缘检测时,会丢失边缘细节信息,产生伪边缘,因此,论文提出在进行Canny算子边缘检测前,首先使用二维最大类间方差算法将图像分为四幅子图,然后对各子图分别进行小波阈值去噪,对去噪图像进行图像融合后,使用Canny算子检测边缘,最后进行边缘连接,得到连续边缘线.实验结果表明,该算法可以在去除噪声的同时有效保留边缘细节,达到较好的检测效果.

Image Segmentation Method Based on Wavelet Transform and Canny Operator

An image segmentation method based on wavelet transform and Canny operator is provided in this paper.As using traditional Canny operator conducting edge detection,edge details are lost and false edges are produced.So the method is provided that the two-dimensional otsu method is used to separate the image into 4 subgraphs before Canny operator is used.Secondly,wavelet threshold denoising is used to each subgraph.Then fusion the subgraphs.Thirdly,the Canny operator is used to extract image edge.Finally,the discontinuous edge line is connected and then the continuous edge line is got.The experimental results show that this method can remove noise while retain edge details effectively and produce good segmentation results.