华南理工大学学报(自然科学版) ›› 2008, Vol. 36 ›› Issue (9): 113-116,121.

• 电子、通信与自动控制 • 上一篇    下一篇

基于数学形态学的图像边缘检测算法

刘清 林土胜   

  1. 华南理工大学 电子与信息学院, 广东 广州 510640
  • 收稿日期:2007-09-24 修回日期:2008-01-07 出版日期:2008-09-25 发布日期:2008-09-25
  • 通信作者: 刘清(1977-),男,博士生,主要从事图像处理、模式识别和图像检索研究. E-mail:liu.qing3@mail.scut.edu.cn
  • 作者简介:刘清(1977-),男,博士生,主要从事图像处理、模式识别和图像检索研究.
  • 基金资助:

    国家自然科学基金资助项目(60472006);广东省自然科学基金团队研究项目(04205783)

Image Edge Detection Algorithm Based on Mathematical Morphology

Liu Qing  Lin Tu-sheng   

  1. School of Electronic and Information Engineering, South China University of Technology, Guangzhou 510640, Gaangdong, China
  • Received:2007-09-24 Revised:2008-01-07 Online:2008-09-25 Published:2008-09-25
  • Contact: 刘清(1977-),男,博士生,主要从事图像处理、模式识别和图像检索研究. E-mail:liu.qing3@mail.scut.edu.cn
  • About author:刘清(1977-),男,博士生,主要从事图像处理、模式识别和图像检索研究.
  • Supported by:

    国家自然科学基金资助项目(60472006);广东省自然科学基金团队研究项目(04205783)

摘要: 传统的边缘提取算法非常有效但对噪声非常敏感,大多形态学边缘检测算法采用单一结构元素,很难对复杂边界进行有效的处理.文中提出一种基于多结构元素多尺度的数学形态学图像边缘检测算法,先用多尺度结构元素交替顺序形态开一闭平滑图像以去噪,再用多结构元素对不同方向的边缘进行提取,最后将各方向边缘融合得到图像边缘.实验结果表明,文中提出的算法不仅具有很强的抗噪性,而且能有效地提取图像的边缘.

关键词: 边缘检测, 数学形态学, 多结构元素, 多尺度, 二值化

Abstract:

It is known that the traditional edge detection methods are effective but are sensitive to noise, and that most mathematical morphology-based edge detection algorithms are insufficient for complex edge feature due to their single structure element. In order to solve these problems, a novel muhi-structure element multi-scale edge detection algorithm based on mathematical morphology is proposed. In this algorithm, the noise is filtered by an alternative-order morphological open-close filter with a multi-scale element, the edge information in different directions is obtained using multi-structure elements and is further fused to generate the final edge. Experimental results show that the proposed algorithm is insensitive to noise and is efficient in complex edge detection.

Key words: edge detection, mathematical morphology, multi-structure, multi-scale, binarization