基于OpenCV的人脸检测系统设计

Design of Face Detection System Based on OpenCV

人脸检测作为人脸识别系统当中一个独立的模块,构成了人脸识别系统中十分关键的部分。本文基于Open CV视觉库,在adaboost算法的基础上,依靠人类肤色在HSV色彩空间中的特性,使用基于HSV色彩空间的肤色分割与adaboost算法相结合的算法,利用Open CV和Visual Studio 2012搭建实验平台,对上述系统进行设计,并对实验结果进行了分析和比较。实验证明基于肤色分割与adaboost算法相结合的算法与adaboost算法相比,能够降低系统误检率。

As an independent module of face recognition system,face detection is a very important part in face recognition system. The main contents are arranged as follows： we add an algorithm that mixed the color segmentation on HSV color space with adaboost algorithm to improve the performance of face detection system. The experimental platform was built using OpenCV and Visual Studio 2012,and then the above system was designed. Finally,we analyze the experiment results to investigate the system function. The experiment shows that comparing with adaboost algorithm,the new strategy can reduce the false detection rate.