Investigation of gesture recognition based on optical fiber Bragg grating sensors. (15th March 2023)

• Record Type: Journal Article
• Authors: Xiong, Li; Guo, Yongxing; Zhu, Jiajing
• Is Part Of: Measurement. Volume 209(2023)
• Publication Date: 2023-03-15
• Subjects: Gesture detection -- Fiber Bragg grating -- Muscle activity; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2023.112498
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: This paper provide a new attempt by embedding the FBG into the polyimide films for gesture recongnition. This approach has several significant advantages, shuch as good creep resistance, good reproducibility, ease of manufacture and installation, quick response. The FBG sensor was used to capture the signal of muscle activity, then to achieve gesture recongnition. The proposed method can repeatedly and accurately recognize gestures with small standard deviations according to a series of performance evaluation experiments. The proposed method can simultaneously identify the category of different gestures and the gender of the performer according to the research data analysis. Abstract: This paper discusses the investigation of optical fiber Bragg grating (FBG) sensors in gesture recognition. We proposed a new attempt by embedding the FBG into polyimide films, then the FBG sensors are arranged on the upper surface of the muscle along the normal direction of the muscle to recognize gestures through detection of muscle activity. A series of experiments were conducted by the selected eight healthy subjects, to quantitatively studied and comprehensively evaluated the performances of this method. Composite experimental systems with multiple gestures, multiple samples and multiple cycles were established for the investigation of the reproducibility and further verification of feasibility. The results show that the FBG sensors has good reproducibility and no chirp effect of the reflection spectrum profile during gesture recognition. The proposed method can effectively detect six gestures with quick responses. Moreover, this method also has the potential to distinguish muscle deformation with different physiological characteristics.