基于人脸的性别识别研究
发布时间：2018-10-22 10:17
【摘要】：人脸是人类很重要的外貌特征。基于人脸的识别功能,包括性别、种族、年龄、身份信息等等逐渐成为近几年研究的重点。基于人脸的性别识别就是赋予电脑人眼所独有的判定性别的能力,使得电脑能够根据输入的人脸图像,进行判别并输出性别分类结果。性别识别在模式识别、人工智能、计算机视觉、信息安全等方面都有着广泛的发展前景和应用价值。本文总结了性别识别所需要的常见人脸数据库、图像预处理的方法和作用,常用的性别特征提取方法和表示方法,以及性别识别算法。这些算法在识别率、计算时间、可扩展性、以及数据存储量方面各有优势和弊端。然后,在Microsoft Visual Studio 2010平台上,基于OpenCV,用C++语言实现了人脸性别识别系统的设计与编程。准备工作包括建立人脸数据库、图像预处理、人脸检测与定位。后期完成了特征向量提取、人脸性别识别等功能。系统中人脸检测采用基于Haar-like特征的Haar分类器,人脸性别识别功能采用Fisherface算法实现。本文实现的性别识别系统完成了以上功能的实现,满足了一般的基于人脸图像的性别识别的分类需求。
[Abstract]:Human face is an important feature of human appearance. Facial recognition, including gender, race, age, identity information and so on, has gradually become the focus of research in recent years. The gender recognition based on human face is to give the computer the unique ability to judge the gender, so that the computer can distinguish and output the result of sex classification according to the input face image. Gender recognition has a broad prospect and application value in pattern recognition, artificial intelligence, computer vision, information security and so on. This paper summarizes the common face database for gender recognition, the methods and functions of image preprocessing, the commonly used methods of gender feature extraction and representation, and the algorithm of gender recognition. These algorithms have advantages and disadvantages in recognition rate, computing time, scalability and data storage. Then, on the platform of Microsoft Visual Studio 2010, the design and programming of face gender recognition system based on OpenCV, and C language are realized. The preparation includes the establishment of face database, image preprocessing, face detection and location. The feature vector extraction and face sex recognition are completed in the later stage. In the system, Haar classifier based on Haar-like feature is used for face detection, and Fisherface algorithm is used to realize the function of face sex recognition. The gender recognition system implemented in this paper has completed the above functions, which meets the general needs of gender recognition based on face image classification.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41
[Abstract]:Human face is an important feature of human appearance. Facial recognition, including gender, race, age, identity information and so on, has gradually become the focus of research in recent years. The gender recognition based on human face is to give the computer the unique ability to judge the gender, so that the computer can distinguish and output the result of sex classification according to the input face image. Gender recognition has a broad prospect and application value in pattern recognition, artificial intelligence, computer vision, information security and so on. This paper summarizes the common face database for gender recognition, the methods and functions of image preprocessing, the commonly used methods of gender feature extraction and representation, and the algorithm of gender recognition. These algorithms have advantages and disadvantages in recognition rate, computing time, scalability and data storage. Then, on the platform of Microsoft Visual Studio 2010, the design and programming of face gender recognition system based on OpenCV, and C language are realized. The preparation includes the establishment of face database, image preprocessing, face detection and location. The feature vector extraction and face sex recognition are completed in the later stage. In the system, Haar classifier based on Haar-like feature is used for face detection, and Fisherface algorithm is used to realize the function of face sex recognition. The gender recognition system implemented in this paper has completed the above functions, which meets the general needs of gender recognition based on face image classification.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41
【参考文献】
相关期刊论文 前7条
1 李建军;;基于双中心羽化WPCA的人脸识别算法研究[J];信息通信;2013年06期
2 付俊辉;杨献峰;;基于人脸图像的性别识别建模与仿真研究[J];计算机仿真;2012年05期
3 李昆仑;张p，
本文编号：2286890
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2286890.html
