SVD与KFDA相结合人脸识别-matlab-毕业论文.doc
文本预览下载声明
目录
摘要 I
Abstract II
1 绪论 1
1.1 人脸识别技术的历史发展 1
1.2人脸识别的研究内容 2
1.3人脸识别研究的意义 3
1.4本文的主要研究内容和安排 4
2. 人脸识别算法原理 5
2.1奇异值方法(SVD) 5
2.2 主分量分析(PCA)方法 6
2.3 Fihser线性鉴别分析 7
2.4 SVD与KFDA相结合人脸识别 9
2.4.1 核Fishe判别分析(KFDA) 10
2.4.2 SVD与KFDA的融合 11
3 实验结果与分析 13
3.1 ORL人脸库实验 13
3.2 CAS-PEAL人脸库实验 15
3.3 结果分析 18
4.总结 19
参考文献 20
致谢 22
附录1 23
摘要
目前于统计特征的线性方法在人脸识别中发展的比较成熟,但是由于人脸识别涉及光照、表情、姿态等问题,线性方法在实际应用中表现的远远不够。因此,将线性方法拓展到非线性领域以提高识别率是一个极待解决的问题。
本论文主要研究了奇异值分解和核Fisher判别分析相结合的方法,将线性Fisher判别算法拓展到非线性方法。既在进行非线性映射之前,首先利用奇异值分解(SVD),提取所得到的奇异值矩阵左上角区域的值作为KFDA的输入空间,再进行进行核Fisher判别分析(KFDA)。在ORL和CAS-PEAL标准人脸库的试验表明,与经典的线性子空间识别方法以及核Fisher鉴别分析(KFDA),它具更高的识别率,识别速度也比较快。
关键词: 人脸识别;奇异值分解 ;核Fisher鉴别分析
Abstract
Face recognition is an important branch of biologic feature identification.Because of its advantages comparing to other biologic features,considerable attention has been paid to face recognition.Due to the importance of human being in the multimedia information,the recognition based on man’s biometrics information is one of the important topics in compute rvision and patten recognition in past 20 years. Many approaches to face recognition problem have been devised,from the early geometry based methods to statistiec based methods.
A fusion of SVD and KFDA for face recognition is developed. The algorithm includes two stages :firstly ,all the train sample are projected into the matrix which come from the SVD of the standar face image.Then the left-top information of projecting coefficient matrix is extracted as the primary feature.Then,used the method of KFDA extracted the finall feature which are used to recognition.ORL and CAS-PEAL database are used to test,the experimental results show the method is effective than many other method such as PCA,LDA,KFDA.
Keywords: face recognition ; singular value decomposition ;Kernel Fisher discriminant Analysis
1 绪论
随着社会的发展以及技术的进步,尤其是最近十年内计算机的软硬件性能的飞速提升,以及社会各方面对快速高效的自动身份验证的要求日益迫切,生物识别技术在科研领域取得了极大的重视和发展。由于生物特征是人的内在属性,具有很强的自身稳定性和个体差异性,因此是身份验证的最理想依据。在人与人的接触中,人脸所包含的视觉
显示全部