利用多特征融合的运动员人体姿势识别算法  

Player’s posture recognition algorithm based on multi-feature fusion

作　　者：刘帅[1] LIU Shuai(Department of Computer Science,Shangqiu Vocational and Technical College,Shangqiu 476100,Henan Province,China)

机构地区：[1]商丘职业技术学院计算机系

出　　处：《信息技术》2019年第8期17-19,23,共4页Information Technology

基　　金：河南省科技厅软科学研究计划项目(142400411213)

摘　　要：针对现有姿势识别算法不能全面反映运动员运动姿势的动态特性问题,文中提出了一种多特征融合的运动员姿势识别算法。首先,通过光学图像采集器采集运动姿势图像,然后将采集到的图像进行灰度变换提高图像质量,进一步基于阴影消除技术和帧间差分法获得身体轮廓和运动姿势区域,最后,基于Radon变换和离散小波变换提取运动姿势区域和身体轮廓,并通过将两种互补性的特征融合实现最终的姿势识别。实验表明,提出的方法识别精度高于其他的较新方法。For the current gesture recognition algorithm cannot fully reflect the dynamic characteristics of the tennis player’s movement posture,a player gesture recognition algorithm based on optical image collector is proposed.Firstly,an athletic pose image is collected by an optical image collector.Then the acquired image is grayscale transformed to improve the image quality,and the body contour and the motion posture region are further obtained based on the shadow elimination technique and the inter-frame difference method.Finally,the motion pose regions and body contours are extracted based on Radon transform and discrete wavelet transform,and the final gesture recognition is achieved by fusing the two complementary features.Experiments show that the proposed method has higher recognition accuracy than traditional methods.

关 键 词：姿势识别 特征提取 帧间差分法 RADON变换 离散小波变换 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]