基于HOG特征与子区域模糊融合的人耳识别研究  

STUDY ON EAR RECOGNITION BY USING HISTOGRAM OF ORIENTED GRADIENT FEATURES AND SUB-REGION FUZZY FUSION

作　　者：封筠[1] 梁晓霞[1] 穆志纯[2] 

机构地区：[1]石家庄铁道大学信息科学与技术学院,河北石家庄050043 [2]北京科技大学信息工程学院,北京100083

出　　处：《计算机应用与软件》2012年第4期79-82,共4页Computer Applications and Software

基　　金：国家自然科学基金面上项目(60573058);河北省科技支撑计划重点项目(10213516D);河北省自然科学基金项目(F2010001106)

摘　　要：作为一种新兴的生物特征识别技术,人耳识别具有其自身独特优势。利用局部特征信息,研究一类新型的基于梯度方向直方图的人耳身份识别方法,提出一种基于梯度方向直方图与子区域模糊融合相结合的人耳识别方案。将人耳图像划分为不同子区域,分别提取各子区域梯度方向直方图特征,引入模糊隶属度匹配融合策略,获取最终的分类结果。与多种方法的对比实验表明,基于梯度方向直方图的特征提取方法具有高识别性能,针对USTB人耳图像库3的测试实验,可达到99.75%的识别率。As an emerging biometric identification technology,ear recognition has the unique advantages of its own.In this paper,a kind of novel ear recognition approach based on histogram of oriented gradient is studied by using the local feature information.An ear recognition scheme based on the combination of histogram of oriented gradient and sub-region fuzzy fusion is presented.The ear image is divided into a number of sub-regions,the features of histogram of oriented gradient in different sub-regions are extracted separately,and the fuzzy membership matching fusion strategy is introduced to obtain the final classification label.Comparative experiments with other methods indicate that the feature extraction method based on histogram of oriented gradient has higher recognition performance.Especially,99.75% recognition rate is obtained while testing on USTB ear image set 3.

关 键 词：人耳识别 梯度方向直方图 子区域模糊融合 K近邻法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]