摘要
针对Android系统自带的人脸检测算法不能精确地检测人脸,尤其是带眼镜后,根本无法检测到人脸。本文研究了一种基于Android系统下的AdaBoost人脸检测算法。首先介绍了Android平台下的人脸检测体系结构,然后对AdaBoost人脸检测模块,包括特征值与特征值的计算、AdaBoost分类器、开发环境搭建分别进行了说明。最后通过样本创建,以及训练好的分类器进行人脸检测。实验结果表明:由于充分利用AdaBoost人脸检测方法实时性比较强、检测率高,该方法完全满足Android平台下人脸检测的需要。
Becacuse Android's own face detection algorithm cannot accurately detect the human face, especially wearing glasses can not detect the human face. In this paper, we study a kind of AdaBoost face detection algorithm based on Android system. Firstly, introduce the Android platform of face detection system structure, then account for the AdaBoost face detection module, including eigenvalue , eigenvalue calculation, AdaBoost classifier and set up the development environment. Finally create the sample and the trained classifier for face detection. The experimental results show that: due to take full advantage of AdaBoost face detection method is more real-time, high detection rate, the method is fully meet the needs of the Android platform face detection.
出处
《电子设计工程》
2014年第8期126-130,共5页
Electronic Design Engineering
基金
国家自然科学基金项目(61201392)
关键词
人脸检测
分类器
样本创建
face detection
Android
AdaBoost
classifier
create sample