AdaBoost基于几何特征的人脸图像检测研究.docVIP

AdaBoost基于几何特征的人脸图像检测研究 摘要人脸检测是人脸信息处理分析中一项十分关键技术，它最初主要来源于人脸识别，是自动人脸识别系统中的一个重要环节，已成为计算机视觉和模式识别等领域内一项十分活跃而又热门的研究型课题，在视频检测与跟踪、基于内容检索、表情识别、视频编码、出入境安全检查、视觉监测等方面都有着非常重要的作用和实际应用价值。 由于目前人脸检测正脸的检测效果比较好，而对侧脸的检测效果就比较差，因此本课题将重心放在了对侧面人脸的检测上，当然，这是在能够检测正脸的基础上。本课题主要是通过检测人眼、鼻子和嘴巴这些特征以及这些特征所独具的几何分布来检测人脸。使用AdaBoost算法训练各特征的分类器来检测人脸五官，对人脸进行统计分析并设计多种人脸特征几何分布的模板，从而实现一个能检测图像中人脸尤其是侧脸的检测系统。12664 关键词侧脸检测AdaBoost Opencv特征检测特征组合侧脸模板 毕业设计说明书（论文）外文摘要 TitleThe Design and Implementation of The detection of Face image Based on The geometric characteristics Abstract Face detection is a crucial technology in face information processing and analysis. It initially comes from face recognition. It is an important part of automatic face recognition system and it has become a very active and popular research topic in the field of computer vision and pattern recognition. Face detection has a very important role and practical application value in aspects of video detection and tracking, content-based retrieval,face recognition ,video coding ,immigration and security checks along with visual monitoring. Currently, the result of the front face detection is relatively good. However, the detection results of the side faces are rather poor. This subject will focus on detecting side of the face. Of course, it is based on being able to detect the front face. The topic detects the human face in the picture mainly through the detection of the characteristics as the human eyes, nose and mouse and the unique geometric distribution of these characteristics. We use the AdaBoost algorithm to train the classifier of features and then use these classifiers to detect the facial features. We conducted a statistical analysis of human faces and designed a variety of facial feature geometric distribution templates. With these templates, wetry to achieve a face detection system ,which can detect the human face from the image, especially the side faces. 3.1初步分析13 3.2实现细节及说明18 3.2.1 核心函数介绍18 3.2.2 检测的实现细节和说明19 4.实验结果与分析21 4.