Image edge detection method based on improved LOG operator

GUAN Liming, LI Lei, LIN Jian
(College of Mechanical Engineering, Hangzhou Dianzi University, Hangzhou 310018, China)

Abstract: Aiming at improving the uncertainty of Gauss coefficient in the traditional edge detection algorithm based on LOG operator, an improved LOG operator was presented. Polar parameter θ was introduced in the proposed algorithm. The alue of θ and the edge point could be obtained by calculating first derivative of θ. The results indicate that the improved LOG operator could get better result than traditional LOG operator, it can get more marginal information, less inveracious edge. While compared with Prewitt operator, Canny operator and other algorithms, it has obvious advantages.
Key words: edge detection; LOG operator; Gauss coefficients