计算机应用与软件2011，Vol.28Issue(12)：54-56,3.
一种基于矢量PCNN的彩色图像自动分割算法
A COLOUR IMAGE AUTOMATIC SEGMENTATION ALGORITHM BASED ON VPCNN
摘要
Abstract
In order to execute colour image automatic segmentation, in the colour image RGB space, the traditional pulse couple neural networks (PCNN) is improved and popularised. A colour image automatic segmentation algorithm based on exponent entropy vector pulse coupling neural networks (VPCNN) is put forward. While taking into account VPCNN linking vector neurons' dynamic temporal-spatial similarities , the method takes advantage of improving the information comparison between exponent dynamic threshold vector and neuron interior activity item vectors to confirm how to segment image target and background area, combine maximum exponent entropy fact to realize colour image automatic segmentation and make comparison against maximum Shannon entropy principle VPCNN segmentation method. Experimental result shows that the algorithm described in the article is precise at image segmentation, strong at adaptivity, and complete at preserving colour image edges, details and so on.关键词
彩色图像处理/图像分割/矢量脉冲耦合神经网络(VPCNN)/指数熵Key words
Colour image processing/Image segmentation/Vector pulse coupling neural networks (VPCNN)/Exponent entropy分类
信息技术与安全科学引用本文复制引用
刘勍,马小姝,张利军,马义德,董忠..一种基于矢量PCNN的彩色图像自动分割算法[J].计算机应用与软件,2011,28(12):54-56,3.基金项目
国家自然科学基金(60872109) （60872109）
甘肃省自然科学基金计划项目(1010RJZE028) （1010RJZE028）
天水师范学院"青蓝"人才工程基金项目及天水师范学院中青年教师科研项目(TSA1001). （TSA1001）
