首页 | 本学科首页   官方微博 | 高级检索  
     

一种基于视皮层神经元模型的彩色图像签名算法
引用本文:寇光杰,马云艳,岳峻,邹海林.一种基于视皮层神经元模型的彩色图像签名算法[J].计算机科学,2015,42(Z11):189-191, 212.
作者姓名:寇光杰  马云艳  岳峻  邹海林
作者单位:鲁东大学信息与电气工程学院 烟台264025,鲁东大学数学与统计学院 烟台264025,鲁东大学信息与电气工程学院 烟台264025,鲁东大学信息与电气工程学院 烟台264025
基金项目:本文受国家自然科学基金项目(61170161,2,61100115),山东省科技发展计划项目(2012YD01056),山东省自然科学基金项目(ZR2012FM008),鲁东大学博士基金项目(LY201222,LY2013001)资助
摘    要:针对彩色图像签名问题进行了研究,基于哺乳动物视皮层神经元工作原理,提出了一种三通道脉冲发放皮层模型TSCM(Triple-Channel Spiking Cortical Model),实现了对彩色图像不变性特征的有效提取。TSCM模型不但具有常规脉冲耦合神经网络签名算法的平移、旋转、缩放不变性,而且算法更加简洁高效,对于噪声影响具有更强的鲁棒性。实验仿真结果验证了该算法的有效性。

关 键 词:脉冲发放皮层模型  脉冲耦合神经网络  彩色图像签名  视皮层神经元

Color Image Signature Algorithm Based on Visual Cortex Neuron Model
KOU Guang-jie,MA Yun-yan,YUE Jun and ZOU Hai-lin.Color Image Signature Algorithm Based on Visual Cortex Neuron Model[J].Computer Science,2015,42(Z11):189-191, 212.
Authors:KOU Guang-jie  MA Yun-yan  YUE Jun and ZOU Hai-lin
Affiliation:School of Information and Electrical Engineering,Ludong University,Yantai 264025,China,School of Mathematics and Statistics,Ludong University,Yantai 264025,China,School of Information and Electrical Engineering,Ludong University,Yantai 264025,China and School of Information and Electrical Engineering,Ludong University,Yantai 264025,China
Abstract:After studying the color image signature problem,a triple-channel spiking cortical model(TSCM)was proposed based on the working principle of mammalian visual cortex neuron model.The invariant characteristics of color image can be extracted effectively by TSCM.On the one hand,TSCM has the properties of ordinary pulse coupled neural network,such as invariances of translation,rotation,and scale.On the other hand,TSCM has more robust characteristics when it faces with the noise.At the same time,the algorithm is more compact and efficient.In the end,the effectiveness of TSCM was proved by the results of experiment and simulation.
Keywords:Spiking cortical model  Pulse coupled neural network  Color image signature  Visual cortex neuron
点击此处可从《计算机科学》浏览原始摘要信息
点击此处可从《计算机科学》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号