| 融合整体与局部特征的人脸识别算法 |
| |
| 引用本文： | 赵建民,唐金良,徐慧英,朱信忠.融合整体与局部特征的人脸识别算法[J].计算机工程与科学,2009,31(7). |
| |
| 作者姓名： | 赵建民 唐金良 徐慧英 朱信忠 |
| |
| 作者单位： | 浙江师范大学数理与信息工程学院,浙江,金华,321004 |
| |
| 基金项目： | 网家自然科学基金资助项目，浙江省科技计划资助项目 |
| |
| 摘 要： | 针对光照、表情、遮挡物等因素的影响,本文提出了一种融合整体和局部特征的人脸识别算法。首先,通过KPCA提取人脸的全局特征;然后,采用简单的图像划分方法将人脸划分成均匀小块,并用KPCA方法分别提取各块特征;最后,基于D-S证据理论的原理对整体与局部特征进行决策级融合得出最终识别结果。实验表明,该算法适应性强,识别率高。
|
| 关 键 词： | 核主成分分析 D-S证据理论 局部特征 整体特征 特征融合 |
A Face Recognition Algorithm Based on Global and Local Feature Fusion |
| |
| ZHAO Jian-min,TANG Jin-liang,XU Hui-ying,ZHU Xin-zhong.A Face Recognition Algorithm Based on Global and Local Feature Fusion[J].Computer Engineering & Science,2009,31(7). |
| |
| Authors: | ZHAO Jian-min TANG Jin-liang XU Hui-ying ZHU Xin-zhong |
| |
| Affiliation: | School of Mathematics;Physics and Information Engineering;Zhejiang Normal University;Jinhua 321004;China |
| |
| Abstract: | A novel algorithm of face recognition based on the fusion of global and local features is proposed to decrease the influence of illumination,facial expressions,cloak and other factors over the recognition rate. Firstly,the global feature is extracted by KPCA,then the face image divided simply into even parts and the feature of every sub-block extracted by KPCA.Finally,the final results by the fusion of global and local features can be concluded based on the principle of the D-S evidence theory.The experimen... |
| |
| Keywords: | KPCA D-S evidence theory local feature global feature feature fusion |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
