基于神经元振荡子网络的视觉场景分割  

Visual scene segmentation based on neural oscillator network

作　　者：刘晓杰[1] 晋卓 宋占伟[3] 张旻[4] 沈琳[1] 

机构地区：[1]江苏理工学院电气信息工程学院,江苏常州213001 [2]吉林省通信管理局,长春130001 [3]吉林大学电子科学与工程学院,长春130012 [4]江苏理工学院计算机工程学院,江苏常州213001

出　　处：《吉林大学学报（工学版）》2013年第S1期244-248,共5页Journal of Jilin University:Engineering and Technology Edition

基　　金：江苏省科技计划项目(BK2012584);江苏省高校自然科学研究项目(12KJD510005);常州市科技计划项目(CE20110034)

摘　　要：为了在图像分割中使视觉场景分割成具有一致性的模块,本文根据神经和脑科学的神经元振荡相关理论,建立了神经元振荡子网络。利用相同物体对应刺激的神经元趋于同步,而不同物体所对应的神经元趋于异步的神经机制,实现视觉场景的分割。通过计算机模拟实验表明,通过合理调整参数,该方法可得到较优的场景分割效果。To solve the consistency problem of divided visual scenes,the neural oscillator network was established by the neural oscillation correlation theory in neural and brain science.The visual scene segmentation was implemented based on the neural mechanism,which was under the corresponding stimulation,the neurons became synchronized between same objects,and the neurons became synchronized between different objects.The experiment results by computer simulations show that the better visual scene segmentation is achieved by the system parameters adjusted.This method is superior to the traditional scene segmentation method in the implementation.

关 键 词：神经元振荡子网络 场景分割 振荡相关 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]