摘要
针对脉冲耦合神经网络(PCNN)模型在数字图像处理中存在的参数庞大、自适应设置和迭代终止条件判定困难等问题,提出了一种基于图像复杂度的PCNN边缘检测新算法。该算法从PCNN数学模型出发,在保留模型同步脉冲发放特性和捕获特性的基础上,对模型进行了数学形式上的简化,减少了模型中参数的数量,同时结合图像复杂度提出参数自适应设置方法。经过实验论证,结果表明该算法能获得完整的图像边缘轮廓和细节,实现PCNN模型实用化、智能化。
In view of large group of parameters and difficulty in adaptive setting and termination decision in digital image based on pulse coupled neural network( PCNN) model,a novel algorithm for edge detection of PCNN based on image complexity was proposed. The form of mathematical equation has been simplified and the number of parameters has been reduced based on the characteristic of synchronization pulse emission and capture feature from mathematical model of PCNN. Meanwhile,the method of adaptive setting has been presented based on image complexity. The experimental results show that the novel algorithm could achieve complete image outline and the details for practicality and intelligence of PCNN model.
出处
《弹箭与制导学报》
CSCD
北大核心
2015年第4期154-158,共5页
Journal of Projectiles,Rockets,Missiles and Guidance
基金
国家自然科学基金(51075395)
国家863计划课题(2013AA040604)
博士后基金(133798)资助