基于交叉视觉皮质模型二值图像形态学算法  

A Binary Image Morphological Processing Algorithm Based on Intersecting Cortical Model

作　　者：徐志平[1] 钟亦平[1] 张世永[1] 

机构地区：[1]复旦大学计算机与信息技术系,上海200433

出　　处：《计算机辅助设计与图形学学报》2007年第8期991-994,1001,共5页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家"十一五"科技支撑计划(2006BAH01B04)

摘　　要：交叉视觉皮质模型（ICM）的神经元受到二值图像中背景信息的激发并将点火信息在神经网络中并行传播,根据ICM具有相似神经元同步激发从而产生脉冲输出的特性,研究和实现了图像形态学的腐蚀、膨胀和骨架化.实验结果表明,利用ICM对二值图像处理的结果等同于数学形态学中的处理结果,且较传统数学形态学中的处理效率平均快3～5倍.The neurons of intersecting cortical model （ICM） are fired by the background information in the binary image and the fired message is broadcasted in parallel cross the neural network. We apply ICM to binary image erosion, dilation, and skeleton. In the mean time, the results generated from the ICM on binary image are equivalent to those generated from the classical mathematical morphology, and the computational efficiency of ICM based method is 3 to 5 times faster than that of the traditional mathematical morphological method.

关 键 词：交叉视觉皮质模型 图像形态学 图像处理 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]