A novel approach for edge detection based on the theory of electrostatic field

作者:Wang, Zhong-Ren*; Quan, Yan-Ming
来源:2007 International Symposium on Intelligent Signal Processing and Communications Systems, ISPACS 2007, Xiamen, 2007-11-28 To 2007-12-01.
DOI:10.1109/ISPACS.2007.4445873

摘要

This paper proposed a novel, simple and effective edge detection algorithm basing on the Law of Coulomb. The algorithm assumes that each pixel is a point electric charge which exerts forces on other neighboring pixels and receives forces from the neighboring pixels. The attracting or excluding forces can be calculated by the law of Coulomb. The vector sum of all electrostatic forces is used to compute the magnitude and the direction of signal variation. Edges are characterized by high magnitude of electrostatic forces along a particular direction and can therefore be detected. The proposed algorithm was tested and compared with other common methods such as Prewitt, LOG, and Canny using Lena image, with and without the corruption of salt &pepper noise. Results show that the proposed edge detector is more robust under noisy condition. In addition, the algorithm can work at any desired scale through regulating the weighting coefficient.