A Compact Wearable Data Glove Based on Flexible Beam Sensors
Tao Yu, Junjie Luo, Yuanqing Gong, Hao Wang, Genliang Chen · 2023
Collecting finger joint angle by data gloves has broad application prospects in interpersonal interaction and remote operation. Typical conventional approaches use heavy rigid mechanism, fatigable silicone frame or complex vision algorithm. In this paper, a new lightweight wearable data glove device based on resistive flexible beam deflection sensors is proposed. 5 resistive strain sensors are attached to the back of a carbon fiber beam, which is fixed with the finger at its tip and root. The beam is modeled as discretized segments linked by flexible joints based on the Principle Axes Decomposition theory. A linear system of geometric and force equations is later established through the Produce-of-Exponential (POE) formula. Afterwards, 5 local curvatures of the beam are derived form linear relations between sensors’ resistance and bending curvatures, then sent to the linear system to reconstruct beam posture. The finger joint postures is finally obtained through the Pythagorean theorem. Calibration experiments are conducted to determine the linear coefficient between resistance and curvature of the sensors. Experimental results show that the best accuracy of reconstructing finger joint angles for an plastic index finger model are 5.8mm, 7.5° for endpoint positions and 7.3° for joint angles. Through wearing tests, the entire measuring device doesn’t restrict the movement of fingers. These results indicate that the flexible beam sensor is efficient and convenient for collecting finger joint angles.