关键词: 图像分割, 脉冲耦合神经网络, 布谷鸟算法, 萤火虫算法, 粒子群算法, 遗传算法

Abstract: Abstract: Concern the problem that the segmentation result of the maximum interclass variance criterion is not enough, the problem of real-time difference is poor and the number of iterations in the Pulse Coupled Neural Network model is difficult to determine, a group of intelligent optimization OTSU-H and Pulse Coupled Neural Network automatic image segmentation algorithm was proposed. The proposed algorithm is composed of three phases. Firstly, the gray information and related information of the image are used to fuse the redundancy, competition and complementarity of the image information effectively, at the same time, the two-dimensional and three-dimensional observation space were constructed. The fast recursive algorithm of OTSU-H criterion was proposed; Secondly, the fitness function of the fast recursive algorithm was used as the fitness function of the four groups of Cuckoo Search, Firefly Algorithm, Particle Swarm Optimization and Genetic Algorithm; Finally, the optimized OTSU-H was introduced into the Pulse Coupled Neural Network model to automatically acquire the number of iterations. The experimental results show that compared with the original OTSU, the maximum entropy criterion and the six latest image segmentation algorithms proposed by the researchers in the past three years The proposed algorithm has better image segmentation effect, while reducing the computational complexity, saving the computer's storage space, and has strong anti-noise ability. In addition, the proposed algorithm has less time loss and no need to train the characteristics, which makes the application of a wide range of algorithms.

Key words: image segmentation, Pulse Coupled Neural Network（PCNN）, Cuckoo Search（CS）, Firefly Algorithm(FA), Particle Swarm Optimization(PSO), Genetic Algorithm(GA)