多变异自适应遗传算法特征筛选的人脸检测  

Haar features selection of face detection based on multiple mutation adaptive genetic algorithm

作　　者：徐柏科 李春贵[2] 阳树洪[2] 徐浩[1] 付行 

机构地区：[1]广西科技大学电气与信息工程学院,广西柳州545006 [2]广西科技大学计算机学院,广西柳州545006

出　　处：《广西工学院学报》2013年第3期73-77,共5页Journal of Guangxi University of Technology

基　　金：广西自然科学基金(2011gxnsfa018162)资助

摘　　要：通过研究基于AdaBoost算法的人脸检测,针对提取的Haar特征维数过高且有冗余,耗时过大,严重影响了人脸检测速度等问题,提出了基于多变异位自适应遗传算法(Multiple Mutation Adaptive Genetic Algorithm,MMAdapGA)的特征选择人脸检测方法.MMAdapGA采用类内类间比作为适应度函数,寻找其最优个体所对应的基因链,然后演化得到全局最优特征,以此降低特征空间维数及冗余度.实验结果表明:检测率(DR)与误检率(FDR)不仅没有下降,而且检测速度得到了很大的提升.Through the study of face detection based on AdaBoost algorithm, in view of the extraction of high dimension, redundancy and being time consuming of Haar features, which seriously impact on the speed of face detection, facial feature selection method was proposed based on more mutation adaptive genetic algorithm （MMAdapGA）. MMAdapGA fitness function is the ratio of intra-class and inter-class. The study first looks for the best individual of the chain of genes, then switches to get the global optimal characteristics so as to reduce the feature space dimension and the degree of redundancy. The experimental results show that, detection rate （DR） and the false detection rate （FDR） do not decline, the detection speed, however, greatly ascend.

关 键 词：人脸检测 HAAR特征 遗传算法 特征选择 类内类间比 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]