Image Segmentation Combining Pulse Coupled Neural Network and Adaptive Glowworm Algorithm

Image segmentation is one of the key steps target recognition. In order to improve accuracy image segmentation, an algorithm combining Pulse Coupled Neural Network(PCNN) and adaptive Glowworm Algorithm(GA) proposed. The retains advantages GA. Introduce moving step size population optimal value as adjustment factors. Enhance ability solve global value, takes weighted sum cross entropy, information entropy compactness fitness function Maintain diversity features improving segmentation. Experimental results show that compared with other algorithms, segmented obtained by this has better visual effect performance best comprehensive performance. For seven gray-scale images in Berkeley dataset, improved 10.85% TDE algorithm, 9.22% GA 22.58% AUTO algorithm.