A Swarm Intelligence Based Medical Image Edge Detection Method with Adaptive Gradient

Edge detection in medical images is a fundamental and crucial step in processing medical images. Due to its advantage of noise insensitivity, the ant colony optimization-based edge detection method is receiving increasing attention. However, existing ACO (ant colony optimization)-based methods adopt a fixed number of neighborhood pixels to compute the gradient in heuristic information for each pixel, and the movement of ants tends to be trapped in local optima and suffer from premature convergence. This causes the loss of important edges and/or the detection of meaningless edges. To address these issues, a novel ACO-based method was proposed in which various numbers of neighborhood pixels were utilized to compute the gradient for each pixel based on the image content around it using a strategy to adaptively and simultaneously determine the perceived radius during the movement of ants. The theoretical analysis and experimental results of this study demonstrated that the proposed method, when compared to highly-rated methods, detected more precise edges.