Ant colony clustering based on cloud model and its application in image segmentation

This paper proposes an ant colony clustering algorithm of image segmentation based on cloud model. Making use of the characteristics of cloud model, it is able to dynamically adjust the value of the pheromone and the pheromone evaporation factor that differ from the traditional ant colony algorithm. Since ant colony algorithm has high ability to deal with local optimization and fuzzy clustering as well as its overall consideration of various factors such as the characteristic of gray and gradient of each pixel and the different membership of each pixel to the object, boundary, background and the noise, the algorithm can segment the image efficiently through the clustering of the pixels with different characteristics. For which followed by transforming the image with watershed algorithm.