Enhanced Image Edge Detection Using Ant Colony Optimization: A Bio-Inspired Meta-Heuristic Approach

Abstract

Image processing plays a pivotal role in modern technological applications, from medical diagnostics to military systems. Among its fundamental tasks, edge detection is critical for object recognition, segmentation and pattern analysis. However, conventional edge detection techniques such as Sobel and Canny often suffer from noise sensitivity, high computational costs and discontinuity in edge localization. This research introduces a robust image edge detection method based on Ant Colony Optimization (ACO), inspired by the foraging behavior and pheromone-based communication of ants. The algorithm simulates the dynamic decision-making of ants to develop a pheromone matrix that effectively identifies intensity changes in 2D images. Experimental evaluations using images like the Taj Mahal, butterfly and rose demonstrate that the proposed ACO method achieves improved Peak Signal-to-Noise Ratio (PSNR) and reduced Mean Square Error (MSE) compared to traditional methods. The study also presents enhancements using adaptive pheromone sensitivity and proposes hybrid strategies involving Bee Colony Optimization for future work. This biologically inspired technique offers a powerful, adaptable and accurate solution for advanced image edge detection challenges.