人体背影序列运动特征的性别识别方法  

Gender recognition of human shadow sequences in complex scenes

作　　者：唐浩[1,2] 李哲媛[1] 刘宏[1] 

机构地区：[1]北京大学深圳研究生院深圳物联网智能感知技术工程实验室,广东深圳518055 [2]湖南师范大学工程与设计学院,湖南长沙410006

出　　处：《华中科技大学学报（自然科学版）》2013年第S1期121-124,共4页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：基金项目国家自然科学基金资助项目(60875050;60675025);国家高技术研究发展计划资助项目(2006AA04Z247);深圳市科技计划资助项目(JCYJ20120614152234873;CXC201104210010A;JCYJ20130331144716089;JCYJ 20130331144631730)

摘　　要：提出一种在复杂场景下基于动态局部二进制模式(VLBP)和三正交平面局部二进制模式(LBP-TOP)的人体背影图片序列运动特征的性别识别方法.首先对图像序列进行预处理,即对图像序列体积块进行分块和三维滤波后,再分别对分块得到的图像序列子体积块用VLBP和LBP-TOP进行运动特征提取.然后用支持向量机对训练样本进行训练和预测得到运动人体的性别.最后研究了分块数量、时间半径和滤波等因素对识别率的影响.研究结果表明:该方法能有效地提取动态人体特征,尤其是在光照条件和地形条件变化较大的复杂场景中效果更好;在自建的数据库上用VLBP特征取得了83.33%的正确率,用LBP-TOP特征取得了94.44%的正确率.A gender recognition algorithm based on VLBP(volume local binary patterns)and LBP-TOP(local binary patterns from three orthogonal panels)of human shadow sequences motion feature in complex scene was proposed.Firstly,VLBP and LBP-TOP were adopted with image sequences preprocessing namely image sequence blocks blocking and three-dimensional filtering to extract dynamic behavior characteristics.Secondly,the support vector machine(SVM)was trained to predict motion human gender with various samples.Finally the influences of blocks,time radius and filtering on recognition accuracy were analyzed.Experiments indicate that this algorithm can extract dynamic features effectively,especially in complex scenes with large illumination variation and background changes.On our database,the SVM with VLBP achieves the recognition accuracy of 83.33%,and the SVM with LBP-TOP achieves 94.44%,respectively.

关 键 词：性别识别 人体背影 图像序列 特征提取 复杂场景 支持向量机 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]