基于adaboost算法的人脸检测.doc

摘要
模式识别中的一个重要分支——人脸识别是当今计算机视觉领域的非常要重的研究方向,同时人脸识别是应用于各个领域的用于身份识别的有效技术手段,在图片或者视频中对人脸的检测是人脸识别的基础,一种高精度的人脸识别技术必须先在图像或者视频中先将人脸定位而后才能够做到有效的人脸识别。所以说人脸检测在人脸识别领域是非常重要的基础研究领域。经过多年来,计算机相关领域的不断发展,人们已经掌握了不少的有关的比较成熟的理论和技术。
在本文中,我们首先介绍有关模式识别的基础知识,模式识别是人脸识别的基础,对模式识别的理解和熟悉有助于我们对人脸识别更加深刻的认识。接着,我们开始回顾人脸识别技术起源、发展历程,这个过程我们可以把它归纳性的划分为三个阶段。
再接着,我们开始介绍各种主要的人脸识别技术,人脸识别的技术主要包括:
几何特征的人脸识别方法;;;;;。
紧接着,我们我们介绍几种主要的人脸识别算法,主要分为:基于人脸特征点的识别算法、基于整幅人脸图像的识别算法、基于模板的识别算法、利用神经网络进行识别的算法。
然后,我们开始介绍本次设计主要使用的人脸检测算法——adaboost算法,以及相关知识——haar矩形特征和积分图。
最后,我们利用opencv对基于adaboost算法的人脸检测技术进行实现。
关键词:人脸识别人脸检测 adaboost haar特征
ABSTRACT
An important branch of pattern recognition - Face recognition is today very weight of the field puter vision research direction, at the same time face recognition is used in various fields of effective techniques for the identification, in pictures or videosface detection is the basis for face recognition, a high precision face recognition technology must be in the image or video first face location before we can achieve effective face recognition. So face detection is a very important field of basic research in the field of face recognition. After years, the continuous development puter-related fields, people have mastered a lot of the more mature theory and technology.
In this article, we first introduce the basics of pattern recognition, pattern recognition is the basis of face recognition to help our face recognition deeper understanding, pattern recognition, understanding and familiar. Next, we began reviewing the origin of face recognition technology, the course of development, a process that we can put it inductive divided into three stages.
Next, we began to introduce a variety of face recognition technology, face recognition technologies include:
Geometric features face recognition; face recognition method based on the characteristics of the face; work face recogniti