基于小波变换与数学形态学的图像边缘检测方法

An Image Edge Detection Method Based on Wavelet Transform and Mathematical Morphology

为得到更加连续完整的边缘图像,提出一种基于B样条小波变换与数学形态学的图像边缘检测方法,在此方法中给出有效重构算法。针对自然图像,利用B样条小波变换作用于图像的高频子图像得到高频边缘图;利用数学形态学检测图像的低频子图像得到图像的低频边缘图;利用文中给出的有效重构的方法将高频边缘图和低频边缘图重构得到最终的边缘图像。实验结果表明,新方法提取出的边缘效果较之前更佳。

To obtain more continuous and complete edge images,the paper proposes an image edge detection method based on B-spline wavelet transform and mathematical morphology,and an effective reconstruction algorithm is given in this method.For natural images,B-spline wavelet transform is applied to the high-frequency sub-images to get high-frequency edge images;mathematical morphology is used to detect the low-frequency sub-images of the images to get the low-frequency edge images;and the high-frequency edge images and low-frequency edge images are reconstructed to get the fnal edge images by the effective reconstruction method given in this paper.Experimental results show that the edge effect obtained by the new method is better than the previous one.