A New Algorithm Based on Multi-scale Order Morphology for Gray Image Edge Detection

In this paper, we present the basic concept of order-morphology and multi-scale structure elements. For gray image segmentation, we develop a new approach for edge detection by combining order-morphology and multi-scale structure elements. Here, the image is convolved with a set of multi-scale structure elements, after then sorted by order morphology, at last all the result images are integrated into one according to how much edge clues it reflects. Compared to traditional Laplacian, Robert and Canny edge detectors, the experiments show that this proceeding produces clearer and more continuous edges, particularly to noised pictures. It obtains good anti-noise capacity due to both order-morphology and multi-scale elements could suppress the influence of noise when detecting edges.