基于人脸识别的电子门锁+程序
毕业论文关键词：嵌入式；OpenCV；人脸识别；Linux；计算机视觉
Abstract This paper introduces a research about the technology of face recognition, which basedon RaspberryPi 2B(4-core CPU and Cortex-A7). Our software platform is Raspbian (Debian-based Linux operating system) and the computer vision library is OpenCV2.4.9.Webuilt a complete embedded platform of computer vision, with the wiringPi library, whichcan operate the GPIO. On the basis of the ARM embedded platform, we studied on how toaccomplish the face recognition. With the great effort, we have resolved how to obtain thevideo sources by cameras or other video capture devices and how to process that. Finally,we completed a control system of electronic lock with the face recognition technologythrough capture, storage, extraction and the process of matching.Our experiments and tests show that the electronic lock based on face recognition canrun stably for a long time. The embedded platform, which we set up, can run the visual library of OpenCV2.4.9 very well. The system can capture the human face accurately and match face quickly, then it can give us a result of matching and control the electronic lock correctly.
Keywords: Embedded Technology；OpenCV；Face Recognition；Linux；ComputerVision
目 录
第一章 绪论1
1.1 研究背景及其意义1
1.3 国内外常见人脸识别的方法-3
1.4 本文的研究内容及组织架构3
第二章 人脸识别原理分析-5
2.1 OpenCV计算机视觉库介绍5
2.2 基于级联分类器的人脸检测6
2.2.1 级联分类器的训练6
2.2.1 级联分类器的检测7
2.3 SURF特征检测与匹配7
2.4 人脸识别技术的性能-10
2.4.1 测量人脸识别的主要性能-10
2.4.1 级联分类器的检测-10
第三章 开发平台介绍与搭建-12
3.1 Raspberry 2B介绍-12
3.1.1 Raspberry 2B配置、功能介绍-12
3.1.2 Raspbian系统-13
3.1.3 GPIO操作-14
3.2 Qt Creator开发平台介绍-16
3.3 开发平台搭建-16
3.3.1 硬件平台搭建-16
3.3.1 软件平台搭建-17
第四章 人脸识别系统设计21
4.1 视频图像提取-22
4.2 人脸检测-23
4.3 人脸样本提取及样本匹配-26
4.3.1 按键值捕捉及判断-27
4.3.2 拍照模式-29
4.3.3 匹配模式-30
4.4 电子门锁控制-33
第五章 实验测试结果与分析-35
5.1 实验测试的人脸库介绍-35
5.2 实验测试及数据记录-36
5.2.1 测试方法一-36
下一篇：PLC船舶损管系统设计+梯形图