Research on the optical image edge detection based on the improved LOG operator

Edge detection has been a hot issue in image processing field for many years. This paper focuses on the LOG operator and its improvement methods. Firstly, a new edge detection template is proposed, which increases the detect direction for the original Laplace template and re-allocation of the weight reasonable. Secondly, this paper proposes an adaptive selection strategy of threshold for edge extraction based on the pixel neighborhood correlation. Experimental results show that the proposed algorithm significantly improved with better detection effect compared to traditional LOG operator and others, and very suitable for practical engineering. Key-Words: Image, Edge detection, LOG operator, Template, Threshold selection, Neighborhood correlation