人脸检测与裁剪系统的设计与实现.doc

四川师范大学本科毕业设计人脸检测与裁剪系统的设计与实现学生姓名院系名称计算机科学学院专业名称软件工程班级学号指导教师完成时间 2016 年5月 10日人脸检测与裁剪系统的设计与实现内容摘要: 本设计旨在开发一个从静态图像中检测人脸及对检测到的人脸图像进行归一化的软件系统,具体内容包括: OpenCV 库函数的调用和内在原理及相关编程环境,掌握使用该库的库函数和接口。在学****过程中了解人脸图像归一化的目和归一化的主要内容与方法。制定了本设计采用的归一化方案,编程实现归一化算法,实现一个基于 OpenCV 的人脸检测和图像归一化系统。 2. 研究了人脸检测原理和算法,学****了 OpenCV 内部封装的人脸人眼检测的算法原理。自己编程实现了一种基于人类皮肤颜色特征的人脸检测算法并对人眼检测方法中的灰度投影法进行了学****和改进。实现了手动定位人眼在图像中的位置,然后根据双眼位置进行人脸图像旋转和裁剪的归一化方法。 、人脸检测的必要性、现状及发展趋势。然后介绍了 OpenCV 人脸检测算法和肤色人脸检测算法的原理和数学模型。之后对本设计实现的人脸检测和裁剪系统的总体设计方案和详细设计方案进行了论述。最后利用 KNN 最近邻分类器,对归一化前后的人脸图像进行了人脸识别实验。实验结果显示,利用本系统可以有效地进行人脸检测和归一化,归一化步骤可以提高人脸识别率。关键词: OpenCV 人脸检测肤色人眼检测人脸归一化 The design and implementation of face detection and cutting system Abstract: This paper aims at develop a face detection and a normaliz ation system. earning The OpenCV (an o puter vision l ibrary ) and the programming environment , master ing the library ’s functions and interface. Get the purpose of normaliz ing and learn some important method and algorithm .A scheme about the implement on this system should be implement the detection system base on OpenCV library , and finish a normalization algorithm by myself. learn some internal encapsulation based on OpenCV by research ing same face detection algorithm of the study I earned the important of face I plan to implement one detection algorithm base on skin color and improv e the human eye Location method named 3. gray-level projection method .The manual positioning of the human eye position pleted, and then rotating and cutting according to the binocular location. 4. First of allI will give a brief describe the history , present situation and the trend of face detection in the future In the paper .Then the principle and mathematical model of two kinds of face detection algorithms based on OpenCV and skin color must be that this paper talk about the overall design of the system design and detail design of this ,use the neares