脉冲耦合神经网络在图像处理中的参数确定

Parameter Determination of Pulse Coupled Neural Network in Image Processing

脉冲耦合神经网络(PCNN)模型可有效地应用于图像处理领域.但目前在PCNN模型理论方面的研究较少,参数的确定仍停留在经验阶段,这很大程度上限制了PCNN模型的发展.本文对PCNN模型进行理论上的推导,特别是模型各参数对PCNN特性的影响,给出了PCNN模型应用于图像处理中各参数确定的准则.在将其应用于眼底图像处理中,取得与人工参数选取相似的效果,表现出较好的鲁棒性.

Pulse coupled neural network （PCNN） can be implemented on image processing effectively. But little researches are about theory analysis in present and the determination of parameters is on the stage of experience, which impedes the development of PCNN model. PCNN model is analysis in theory and the influence of parameters to PCNN is proved in this paper. Finally the rule of parameter determination in image processing is proposed. In ocular fundus image processing, the effect of automation parameter determination is similar to manual parameter determination and the result demonstrate its robustness.