基于数学形态学的图像边缘检测技术

边缘检测通常是用类似于素描图的图像表达出物体的要素和特征。实际图像中,边缘由灰度突变的象素点组成,在数字图像处理和分析中具有重要的作用。本文综合国内外最新文献资料,分析了多种基于数学形态学的边缘检测技术:基于多尺度形态学的边缘检测、基于数学形态学多极平均的图像的边缘检测、基于偏微分方程的形态学的边缘检测、基于均衡化和数学形态学的组合边缘检测、基于坐标逻辑的多结构元图像边缘检测等技术,并综合比较了其优缺点,探讨了其发展方向。

Morphological Edge Detection Technology in Digital Image Processing

The aim of edge detection is usually to use the image that is more similar than sketch picture to express the feature and outline of object. In actual image, edge consists of saltation gray scale pixels .It plays on important role in digital image processing and analysis. This paper synthesizes newest literatures, analyses many kinds of edge detection techniques based on mathematical morphology: multi-scale morphologic edge detection , multi-grade mean morphologic edge detection, morphological edge detector based on partial differential equations , combined edge detecting algorithm based on equalization and morphology , edge detector using multi-structure element based on coordinate (logic,)and so on. The paper also compares their pluses and minuses synthetically , and probes into its development of edge detection.