Estimation of Hand Motion Based on Forearm Deformation
Estimation of Hand Motion Based on Forearm Deformation
复制标题
DOI：
10.1109/robio.2018.8664823
复制
发表时间：
2018-12
期刊：
影响因子：
--
通讯作者：
Sung-Gwi Cho;M. Yoshikawa;Ming Ding;J. Takamatsu;T. Ogasawara
中科院分区：
文献类型：
--
作者：
Sung-Gwi Cho;M. Yoshikawa;Ming Ding;J. Takamatsu;T. Ogasawara
在各种应用中，手部运动估计的研究起着重要的作用。在这些研究中，基于生物信号的方法越来越受欢迎，因为生物信号只能使用可穿戴设备进行测量。在各种生物信号中，我们重点关注前臂变形。前臂变形提供了表层和深层肌肉、肌腱和骨骼的运动信息。先前基于前臂变形的研究仅试图识别手部运动/形状的类型或估计单个关节角度。在这项研究中，我们提出了一种基于距离传感器阵列测量前臂变形的多关节角度手部运动估计方法。关节角回归采用支持向量回归。所述距离传感器阵列具有10个距离传感器单元和可调整为用户前臂大小的机构。手部运动估计实验结果表明，该方法能正确估计关节角，RMSE为6.6度。
In various applications, studies of hand motion estimation play an important role. In these studies, the methods based on biosignals have gained in popularity because biosignals can be measured using only wearable devices. Among the various biosignals, we focus on forearm deformation. The forearm deformation provides the motion information about the surface and deep layer muscles, tendons, and bones. The previous studies based on the forearm deformation only attempted to recognize types of hand motions/shapes or to estimate a single joint angle. In this study, we propose a hand motion estimation method for multiple joint angles based on the forearm deformation measured with a distance sensor array. A Support Vector Regression is used for the joint angles regression. The distance sensor array has 10 distance sensor units and a mechanism that can be adjusted to the size of the user's forearm. The results of the hand motion estimation experiments show that the proposed method correctly estimated the joint angles with an RMSE of 6.6 degrees.