欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
由于人脸姿态、光照和表情等各方面的显著差别,使人脸识别成为非常具有挑战性的模式识别问题。主成分分析是模式识别技术中经典的特征抽取和降维技术之一。本文把其改进算法二维对称主成分分析应用到人脸识别中。二维对称主成分分析与传统主成分分析和对称主成分分析相比,可以得到更好的适合分类的特征。实验结果表明,二维对称主成分分析不仅实现了降维,而且能取得比传统主成分分析和对称主成分分析更好的识别性能,对ORL标准人脸数据库的正确识别率达到94%以上。
英文摘要：
Because of obvious variations in face pose, lighting, and expression, face recognition becomes a very challenging research topic in pattern recognition. Principal Component Analysis （PCA） is one of the classical methods for feature extraction and dimensional reduction. An improved PCA algorithm with two dimension symmetrical PCA （2DSPCA）, was used in face recognition. The better features suitable for categorization were extracted based on 2DSPCA compared with classical PCA. Experiment results also demonstrate that 2DSPCA is not only good at dimensional reduction, but also achieves better performance than classical PCA and SPCA, its recognition rate for ORL database is higher than 94%.
同期刊论文项目
同项目期刊论文
Almost Orthogonality and Hausdorff Interval Topologies of de Morgan Lattices and Lattice Effect Alge
期刊信息
位置：