对偶树复小波与空域信息的手势识别分类研究  

Research on gesture recognition and classification of dual-tree complex wavelet and spatial information

作　　者：贾鹤鸣[1] 朱传旭 张森[1] 杨泽文 何东旭[2] JIA Heming;ZHU Chuanxu;ZHANG Sen;YANG Zewen;HE Dongxu(College of Mechanical and Electrical Engineering,Northeast Forestry University,Harbin 150040,China;College of Automa-tion,Harbin Engineering University,Harbin 150001,China)

机构地区：[1]东北林业大学机电工程学院,黑龙江哈尔滨150040 [2]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001

出　　处：《智能系统学报》2018年第4期619-624,共6页CAAI Transactions on Intelligent Systems

基　　金：中央高校基本科研业务费专项资金项目(2572014BB03);国家自然科学基金项目(31470714;51609048);黑龙江省研究生教育创新工程项目(JGXM_HLJ_2016014)

摘　　要：为提高手势识别中特征获取的有效性,本文提出空域特征与对偶树复小波变换特征相结合的融合特征,主要包括水平位置、竖直位置、长宽比、矩形度、Hu矩7个分量,及11维空域特征与对偶树复小波变换的16维特征进行融合后得到的27维特征。针对分类器优化算法,提出进行训练样本优选的最优距离–支持向量机(BD-SVM)分类方法。最后的实验结果表明,对"1~9"手势进行测试,当采用径向基核函数时,平均识别精度最高,为90.33%,平均识别时间为0.026 s,说明所提出的方法能够较好地进行静态手势识别,具有较高的训练速度和辨识精度。To improve the validity of features obtained in gesture recognition, in this paper, we propose a fusion featurethat combines spatial and dual-tree complex wavelet transform features. These features mainly include seven compon-ents （horizontal position, vertical position, aspect ratio, rectangular degree, Hu moments, etc.） and 27 dimensional fea-tures, comprising 11 dimensional spatial features and 16 dimensional dual-tree complex wavelet transform features. Weemploy the optimal distance support vector machine （BD-SVM） classification method to optimize training samples forthe classifier optimization algorithm. The experimental results show that, in a test of gestures “1~9” using the RBF ker-nel function, the highest average recognition accuracy is 90.33% and the average recognition time is 0.026 s. These res-ults reveal that the proposed method demonstrates excellent static gesture recognition, a high training speed, and accur-acy in identification.

关 键 词：手势识别 空域特征 对偶树复小波 特征融合 分类器优化 BD-SVM 径向基核函数 静态测试 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]