数学形态滤波的细胞神经网络实现方法

Mathematical Morphology Using Cellular Neural Network

给出了细胞神经网络（ＣＮＮ）在数学形态学中的应用．通过对ＣＮＮ状态方程的修改，可使ＣＮＮ并行、实时地实现数学形态滤波，其实现速度与结构元的大小无关．给出了有关改进ＣＮＮ的稳定性及动态特性分析以及在腐蚀、膨胀及结构开。

An application of cellular neural network (CNN) in mathematical morphology is presented.By modifying its original state equation,CNN is made to be capable of carrying out the morphological operation in real time,in which the processing speed is regardless of the morphological structuring element.The theorems concerning the dynamic range and steady state are given while the computer simulating results are used to illustrate the morphological dilation,erosion,open and close on binarized images.