A novel edge detection approach based on backtracking search optimization algorithm (BSA) clustering

Image edge information is very important in application areas such as machine learning, image processing, stereo vision, object tracking and pattern recognition. Intensity discontinuities or sudden intensity changes in a region are indicative of the edge region in that region. Although there are many approaches to detecting edge, generally intensity discontinuities or sudden intensity changes in a region are described as edge. In this study, we proposed a Backtracking Search (BSA) clustering based edge detection approach for noisy images. Proposed approach has two stages. In first stage, the edge map is calculated using the max-min filter defined in a window. In second stage, edge map is calculated via BSA based clustering with using a cost function.