Consistent approach to edge detection using multiscale fuzzy modeling analysis in the human retina

Today, many widely used image processing algorithms based on human visual system have been developed. In this paper a smart edge detection based on modeling the performance of simple and complex cells and also modeling and multi-scale image processing in the primary visual cortex is presented. A way to adjust the parameters of Gabor filters (mathematical models of simple cells) And the proposed non-linear threshold response are presented in order to Modeling of simple and complex cells. Also, due to multi-scale modeling analysis conducted in the human retina, in the proposed algorithm, all edges of the small and large structures with high precision are detected and localized. Comparing the results of the proposed method for a reliable database with conventional methods shows the higher Performance (about 4-13%) and reliability of the proposed method in the detection and localization of edge.