JIANG Yi, ZHANG Tong, XIONG Zhenzhen
(Jingdezhen Ceramic Institute, Jingdezhen 333403, Jiangxi, China)

Abstract: Clustering has become an important method in investigation of many areas such as image processing, computer vision and pattern recognition. In this paper, a new ceramic image segmentation method is provided based on the k-clustering from theory of pattern recognition according to the complexity of all kinds of ceramic images. The proposed method obtains the optimal threshold by the search of particles.Experimental results show that it can get better segmentation results in single threshold and double segmentation, and the time of obtaining the threshold value is shorter than the compared algorithm.
Key words: image segmentation; particle swarm optimization; clustering; ceramic image