基于形态学和小波变换的图像边缘检测方法

电力机房图像边缘提取是实现电力机房三维重建的重要辅助环节,机房图像边缘提取越精准,三维重建将会更准确.本文提出一种融合小波变换模极大值和多尺度多结构形态学的图像边缘检测算法对电力机房图像进行处理.首先,对原始电力机房图像进行小波分解得到高频图像和低频图像;然后,采用小波变化模极大值算法提取高频图像的边缘信息,多尺度多结构数学形态学算法提取低频图像的边缘信息;最后,通过叠加运算融合高频和低频的边缘信息,得到原始图像的边缘信息.通过仿真实验表明,本文提出的边缘检测算法在抗噪性能、边缘连续性、定位精度上综合实力最强.

An Image Edge Detection Method Based on Morphology and Wavelet Transform

Edge extraction of the image of a power room is an important auxiliary link for 3D reconstruction. The more precise the edge extraction of the image, the more accurate the 3D reconstruction. An edge detection method based on wavelet transform modulus maxima and multi-scale and multi-structure mathematical morphology is proposed in this paper to process the image of a power room. First, the original images are divided into high frequency and low frequency components by wavelet decomposition. Then, high frequency component edge extraction is realized by using wavelet transform modulus maxima, and the low frequency component edge extraction is realized by the multi-scale and multi-structure mathematical morphology algorithm. Finally, the whole edge is obtained through image fusion. A simulation experiment shows that our method has good anti-noise performance, precise edge continuity and positioning accuracy.