|
|||||
|
|
| 最小风险贝叶斯决策的二值化人脸识别算法 | |
| 引用本文： | 曾岳,冯大政,付达杰. 最小风险贝叶斯决策的二值化人脸识别算法[J]. 计算机工程与设计, 2011, 32(10): 3511-3513 |
| 作者姓名： | 曾岳 冯大政 付达杰 |
| 作者单位： | 1. 西安电子科技大学雷达信号国家重点研究所,陕西西安710071;江西财经职业学院信息工程系,江西九江332000 2. 西安电子科技大学雷达信号国家重点研究所,陕西西安,710071 3. 江西财经职业学院信息工程系,江西九江,332000 |
| 基金项目： | 国家自然科学基金项目(60372049); 江西省科技计划青年基金项目(GJJ09412) |
| 摘 要： | 提出了一种最小风险贝叶斯决策的二值化人脸识别算法,该算法通过设定图像灰度级的阈值对图像进行二值化,统计其出现的频率,计算其类条件概率密度,根据图像的相似性估计其损失函数,利用贝叶斯公式求最小风险,最后根据最小风险判断其所属类别。该方法克服了传统贝叶斯方法难求类内和类间协方差矩阵的缺点,简单易用。实验结果表明,该方法具有可行性,比传统的基于代数的人脸识别算法(PCA、LDA和PCA+LDA)的识别率高,并能有效减少相似类的重叠。 |
| 关 键 词： | 贝叶斯 人脸识别 后验概率 二值化 最小风险 |
Face recognition algorithm of binary image by smallest risk Bayesian method |
|
| ZENG Yue,FENG Da-zheng,FU Da-jie. Face recognition algorithm of binary image by smallest risk Bayesian method[J]. Computer Engineering and Design, 2011, 32(10): 3511-3513 | |
| Authors: | ZENG Yue FENG Da-zheng FU Da-jie |
| Affiliation: | ZENG Yue1,2,FENG Da-zheng1,FU Da-jie2(1.State Key Lab of Radar Signal Processing,Xidian University,Xi'an 710071,China,2.Department of Information Engineering,Jiangxi Vocational College of Finance and Economics,Jiujiang 332000,China) |
| Abstract: | A face recognition algorithm of binary image by the smallest risk Bayesian method is presented.Firstly by setting the image gray-level threshold the binary image is gotten,then counting its frequency of the pix and calculating the class conditional probability density,this algorithm gets the smallest risk probability to classify it by the Bayesian formula according to it's the loss function gotten by the Image similarity.This method overcomes the shortcomings of traditional Bayesian approach hard to find wi... |
| Keywords: | Bayesian face recognition posterior probability binarization smallest risk |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |