Key points are not available for this paper at this time.
The development of techniques for monitoring finger movement is becoming increasingly important in areas, such as robotics, virtual reality, and rehabilitation. To date, various techniques have been proposed for tracking hand movements, but the majority suffer from poor accuracy and repeatability. Inspired by the articulated structure of finger joints, we propose a novel 3-D printed optical sensor with a compact hinged configuration for tracking finger flexion. This sensor exploits Malus’ law using the attenuation of light transmitted through crossed polarizers. The sensor consists of a single LED, two pieces of linear polarizing film, and a photodetector that detects the changes in polarized light intensity proportional to the angle of finger flexion. This paper presents the characterization of the proposed optical sensor and compares it with a commonly used commercial bend sensor. Results show that the bend sensor exhibits hysteresis error, low sensitivity at small angles, and significant temporal drift. In contrast, the optical sensor is more accurate (±0.5°) in the measuring range from 0° to 90°, and exhibits high repeatability and stability, as well as a fast dynamic response. Overall, the optical sensor outperforms the commercial bend sensor, and shows excellent potential for monitoring hand movements in real time.
Wang et al. (Thu,) studied this question.