人脸识别算法的实现与对比分析+源代码
毕业论文关键词：PCA；KNN；灰度化
Realization and Comparison of Face Recognition Algorithm
Abstract：Face recognition is an active subject in the field of pattern recognition. With its increasing scope of application, it is more and more important for information security, identification , public security, self-service,etc. But the existing face recognition algorithms and source of data. In this paper, we focus on the popular PCA (Principal Component Analysis) and KNN (k-Nearest Neighbor) algorithms. Through the realization and modification of the two algorithms, the basic principles, the respective recognition methods, the way of data processing, the length of the processing time, the respective strengths and the corresponding limitations are analyzed, the processing speed and recognition accuracy of two algorithms are compared, so as to study their own performance and shortcoming, and look forward to the development of face recognition algorithm.
Key words: PCA; KNN; gray processing
目 录
摘要 1
关键字 1
Abstract 1
Key words 1
1．选题背景 1
1.1 问题的提出 1
1.1.1 人脸识别的概念 2
1.1.2 人脸识别的发展历程 2
1.2 国内外研究状况 2
1.2.1 国外研究状况 2
1.2.2 国内研究状况 2
1.3 研究的目的和内容 3
2. 数据采集及处理过程 3
2.1 ORL人脸库简介 3
2.2 Yale人脸库简介 3
2.3 处理流程 4
3. 算法原理 4
3.1 PCA算法 5
3.1.1 PCA算法的概念 5
3.1.2 PCA算法的原理 5
3.1.3 PCA算法的过程 5
3.1.4 PCA算法的优缺点 6
3.1.5 ZCA的原理 7
3.1.6 SVM（支持向量机）的原理 7
3.2 KNN算法 9
3.2.1 KNN算法的原理 9
3.2.2 KNN算法的三要素 9
3.2.3 KNN算法的过程 9
3.2.4 KNN算法的优缺点 10
4. 数据分析 10
4.1 PCA算法 10
4.1.1 实验数据 10
4.1.2 数据分析 12
4.2 PCA、PCA+SVM、ZCA在的ORL人脸库之间的对比 12
4.2.1 实验数据 12
4.2.2 数据分析 13
4.3 PCA算法在不同的人脸库之间的对比 13
4.3.1 实验数据 13
4.3.2 数据分析 13
4.4 KNN算法在不同的人脸库之间的对比 14
4.4.1 实验数据 14