Department of Mechanical Engineering and Construction, Universitat Jaume I, E12071 Castellón, Spain.
Sensors (Basel). 2021 Feb 4;21(4):1049. doi: 10.3390/s21041049.
Simultaneous measurement of the kinematics of all hand segments is cumbersome due to sensor placement constraints, occlusions, and environmental disturbances. The aim of this study is to reduce the number of sensors required by using kinematic synergies, which are considered the basic building blocks underlying hand motions. Synergies were identified from the public KIN-MUS UJI database (22 subjects, 26 representative daily activities). Ten synergies per subject were extracted as the principal components explaining at least 95% of the total variance of the angles recorded across all tasks. The 220 resulting synergies were clustered, and candidate angles for estimating the remaining angles were obtained from these groups. Different combinations of candidates were tested and the one providing the lowest error was selected, its goodness being evaluated against kinematic data from another dataset (KINE-ADL BE-UJI). Consequently, the original 16 joint angles were reduced to eight: carpometacarpal flexion and abduction of thumb, metacarpophalangeal and interphalangeal flexion of thumb, proximal interphalangeal flexion of index and ring fingers, metacarpophalangeal flexion of ring finger, and palmar arch. Average estimation errors across joints were below 10% of the range of motion of each joint angle for all the activities. Across activities, errors ranged between 3.1% and 16.8%
由于传感器放置的限制、遮挡和环境干扰，同时测量所有手部段的运动学是很麻烦的。本研究的目的是通过使用运动协同来减少所需传感器的数量，运动协同被认为是手部运动的基本构建块。协同作用是从公共 KIN-MUS UJI 数据库（22 个受试者，26 个代表性日常活动）中识别出来的。每个受试者提取 10 个协同作用作为主成分，这些主成分解释了在所有任务中记录的角度的总方差的至少 95%。220 个协同作用被聚类，从这些组中获得了用于估计其余角度的候选角度。测试了不同的候选组合，并选择了提供最低误差的组合，其好坏性是根据另一个数据集（KINE-ADL BE-UJI）的运动学数据来评估的。因此，将原来的 16 个关节角度减少到 8 个：拇指的腕掌关节屈曲和外展、拇指的掌指关节和指间关节屈曲、食指和环指的近端指间关节屈曲、环指的掌指关节屈曲以及掌弓。在所有活动中，关节间的平均估计误差低于每个关节角度运动范围的 10%。在所有活动中，误差范围在 3.1%至 16.8%之间。
