Change detection in remote-sensed data by particle swarm optimized edge detection image segmentation technique

Satellite images help in monitoring change detection as they are the big repository of information. An imperative task from the prospects of land development monitoring, disaster management, resource management, and environment evaluation is change detection. For change detection, segmentation of an image is being performed for locating the areas of interest. Nature-inspired particle swarm optimization is a metaheuristic algorithm that is simple, robust, and makes a fewer number of assumptions for the problem considered. This paper implements a particle swarm optimization (PSO) algorithm in MATLAB environment as edge detection segmentation technique for satellite images, which are being acquired from Google Earth. For qualitative analysis, the results are compared with the conventional edge detector operators such as Sobel, Canny, and Prewitt with the help of entropy values. It has been observed that PSO outperforms the conventional edge detection image segmentation methods, thereby giving better edges and clarity in images for change detection.