An Edge Detection Method Based On Computational Model Of Simple Cells In Primary Visual Cortex

Simple cells in primary visual cortex respond to the local, oriented edge segments within their receptive fields. In this study, we present a new edge detection method based on the computational model of these cells. Firstly, the response of a set of simple cells for a number of different preferred orientations are calculated. Then, the intensity gradient for each pixel is obtained using the linear summation of these responses. Some parameters of simple cell computational model are calculated in such a way that a set of goals (good detection, good localization and only one response to a single edge) achieving for the resulting operator. Considering the properties of medical images, the proposed operator is useful for medical image edge detection. The synthesis and medical images with their associated ground truth edge maps are used to assess performance of the proposed method. The results obtained from the proposed method are found to be better and more stable with respect to the input parameters than those from many well known edge detectors (e.g. Canny edge detector).