A niching particle swarm segmentation of infrared images

Niching techniques play an important role in evolutionary algorithms. Considering the characteristics of the inconspicuous difference between targets and backgrounds and the low contrast in infrared images, a new algorithm based on niching particle swarm optimization is used in the infrared image processing to determine the optimal thresholds in image segmentation. The algorithm uses fuzzy C-mean clustering algorithm, by the optimization of the niching particle swarm optimization object function, the optimal thresholds can be gotten, and the infrared image by use of the thresholds can be segmented. Experiments show that the new algorithm can get the optimal threshold by the maximum entropy.