结合粒子群优化和综合评价的脉冲耦合神经网络图像自动分割  

Automated image segmentation based on pulse coupled neural network with partide swarm optimization and comprehensive evaluation

作　　者：张坤华[1,2] 谭志恒 李斌 ZHANG Kun-hua;TAN Zhi-heng;LI bin(College of Information Engineering,ShenzhenUniversity,Shenzhen 518060,China;Shenzhen key laboratory of media information content security,Shenzhen 518060,China)

机构地区：[1]深圳大学信息工程学院,广东深圳518060 [2]深圳市媒体信息内容安全重点实验室,广东深圳518060

出　　处：《光学精密工程》2018年第4期962-970,共9页Optics and Precision Engineering

基　　金：国家自然科学基金资助项目(No.61572329)

摘　　要：为了解决脉冲耦合神经网络(Pulse Coupled Neural Network,PCNN)在图像分割中多参数设定以及评价准则单一的问题,提出了一种结合粒子群优化算法(Particle Swarm Optimization,PSO)和综合评价准则的PCNN图像自动分割方法。采用单调递增阈值搜索策略的PCNN改进模型,将PSO优化原理与由交叉熵参数,边缘匹配度和噪点控制度共同构成的综合评价相结合,以综合评价作为粒子的适应度函数,自动寻优获取PCNN图像分割模型的目标时间常数,连接系数以及迭代次数n,从而实现全参数自适应的PCNN图像分割。实验结果表明算法在保证PCNN运行效率下对不同类型图像都能进行正确完整的分割并兼顾纹理细节的保留。从实验数据可以看到,本文算法在综合评价和通用综合指标上均优于其他对比算法,综合评价平均优于其他算法10.5%。客观评价结果与视觉主观评价相一致,分割较理想,算法具有较高的鲁棒性。Multi-parameter setting and single segmentation evaluation criterion are the problems in image segmentation based on Pulse Coupled Neural Network(PCNN).Through combining Particle Swarm Optimization(PSO)with comprehensive evaluation criterion,this paper presented an automatic image segmentation algorithm based on PCNN.The improved PCNN model with monotonically increasing threshold search strategy was utilized in this algorithm.The Comprehensive Evaluation Criterion(CEC)obtained by cross-entropy parameter,edge matching degree and noise control degree were proposed as the fitness of particles in PSO,then the parameters of PCNN such as the target time constant,the connection coefficient and the iteration times n were acquired adaptively by updating fitness value of particles.By using these acquired optimum parameters,the image was segmented by the improved PCNN.For different types of images,experimental results show that algorithm proposed can segment image completely and accurately under PCNN operating efficiency,moreover texture details are retained.Compared with other experimental methods,the segmented results obtained by proposed algorithm are superior to that obtained by other algorithms in CEC 10.5%.In addition,the general comprehensive indicators of the segmented results obtained in this research are also optimal.Thus,it can be seen that the objective evaluations are consistent with the visual subjective evaluations,and the algorithm proposed has high robustness.

关 键 词：脉冲耦合神经网络 粒子群优化 综合评价 参数自适应 图像分割 

分 类 号：TP394.1[自动化与计算机技术—计算机应用技术]