Infrared image edge detection algorithm based on sobel and ant colony algorithm

Infrared image edge detection is a key subject of image processing. Although ant colony algorithm (ACA) has performed well in edge detection, the ant still waste much time on non-edge district as the initial position was random set. The paper uses Sobel operator controlling the set of the initial position in AC A, make the ant search the global image from those high-probability pixel, it also improve the detection of tiny edge and whole efficiency of algorithm. The results of the multiple-group experiments indicate that the technique include both Sobel operator and ACA´s advantage, give a better detection of infrared image.