Computer Engineering and Applications ›› 2007, Vol. 43 ›› Issue (21): 70-71.

• 学术探讨 • Previous Articles     Next Articles

Edge detection method based on maximal entropy and morphology

WEI Jun-wei,FANG Min   

  1. School of Computer Science,Xidian University,Xi’an 710071,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-07-21 Published:2007-07-21
  • Contact: WEI Jun-wei


魏军伟,方 敏   

  1. 西安电子科技大学 计算机学院,西安 710071
  • 通讯作者: 魏军伟

Abstract: Through the combination of maximal entropy theory and morphology method, a new edge detection method for gray images is proposed in this paper.The experiment indicates that it can obviously improve the quality of edge detection in this method compared with traditional methods,and the detected edges are all single-pixel line.It remains the image information at most while picks the single-pixel edge quickly and precisely.

Key words: gray image, edge detection, morphology, entropy

摘要: 利用最大熵图像分割理论和形态学方法相结合,提出了一种新的灰度图像的边缘提取方法。以一维最大熵和二维最大熵为例,分别与形态学方法结合进行了实验对比。试验表明,该方法与传统边缘检测方法相比,能够在无需参数输入的前提下自动提取出图像边缘,且检测到的边缘都是单像素线条;可以在尽量保留图像最大信息量的基础上,快速准确地提取出单像素封闭边缘。

关键词: 灰度图像, 边缘检测, 形态学,