模拟生物嗅觉神经系统的混沌神经网络及应用

Application on Imaging Pattern Recognition of A Chaotic Neural Network Simulating Olfactory Neural System

以神经生物学实验结果为基础,根据生物嗅觉神经系统的信息处理机制,Freeman建立了非线性神经网络模型-K系列模型。KIII模型在模拟神经系统方面具有突出的优点,同时也具有一定的模式识别的能力,它的仿生特点代表了一种新型的神经网络模型。因此,KIII模型在解释人脑的认知机制和处理比较复杂的模式识别问题中有着广泛的实际应用价值。该文应用KIII模型在图像模式识别方面的应用做了初步探索,并将该模型应用于具体的简单图像模式识别中,取得了良好的效果。这一探索拓展了KIII模型模式识别应用范围,为该模型在图像模式识别方面的广泛应用建立了基础。

K models, which were built by Prof. Freeman, is a new type of nonlinear neural network. According to information processing mechanism of the olfactory neural system, these models are built on the base of a great deal of neural biology experiments. Among these models, KIII model is built to simulate the olfactory neural system producing EEG, which is extracted from mammal olfactory cortex. At the same time, this model has abilities for pattern recognition. We can find this kind of ability of KIII model by applying it on imaging pattern recognition.