In recent years, surface electromyography (sEMG) signals have been increasingly used in pattern recognition and rehabilitation. In this paper, a real-time hand gesture recognition model using sEMG is proposed. We use an armband to acquire sEMG signals and apply a sliding window approach to segment the data in extracting features. A feedforward artificial neural network (ANN) is founded and trained by the training dataset. A test method is used in which the gesture will be recognized when recognized label times reach the threshold of activation times by the ANN classifier. In the experiment, we collected real sEMG data from twelve subjects and used a set of five gestures from each subject to evaluate our model, with an average recognition rate of 98.7% and an average response time of 227.76 ms, which is only one-third of the gesture time. Therefore, the pattern recognition system might be able to recognize a gesture before the gesture is completed.

No abstract

Recent advances in myoelectric controlled techniques have made the surface electromyogram (sEMG)-based sensing armband a promising candidate for acquiring bioelectric signals in a simple and convenient way. However, inevitable electrode shift as a non-negligible defect commonly causes a trained classifier requiring continuous recalibrations. In this study, a novel hand gesture prediction is firstly proposed; it is robust to electrode shift with arbitrary angle. Unlike real-time recognition which outputs target gestures only after the termination of hand motions, our proposed advanced prediction can provide the same results, even before the completion of signal collection. Moreover, by combining interpolated peak location and preset synchronous gesture, the developed simplified rapid electrode shift detection and correction at random rather than previous fixed angles are realized. Experimental results demonstrate that it is possible to achieve both electrode shift detection with high precision and gesture prediction with high accuracy. This study provides a new insight into electrode shift robustness which brings gesture prediction a step closer to practical applications.

No abstract

This paper presents a theoretical method for separating bending and torsion of shape sensing sensor to improve sensing accuracy during its deformation. We design a kind of shape sensing sensor by encapsulating three fibers on the surface of a flexible rod and forming a triangular FBG sensors array. According to the configuration of FBG sensors array, we derive the relationship between bending curvature and bending strain, and set up a function about the packaging angle of FBG sensor and strain induced by torsion under different twist angles. Combined with the influence of bending and torsion on strain, we establish a nonlinear matrix equation resolving three unknown parameters including maximum strain, bending direction and wavelength shift induced by torsion and temperature. The three parameters are sufficient to separate bending and torsion, and acquire two scalar functions including curvature and torsion, which could describe 3D shape of rod according to Frenet-Serret formulas. Experimental results show that the relative average error of measurement about maximum strain, bending direction is respectively 2.65% and 0.86% when shape-sensing sensor is bent into an arc with a radius of 260 mm. The separating method also applied to 2D shape and 3D shape of reconstruction, and the absolute spatial position maximum error is respectively 3.79mm and 11.10mm when shape-sensing sensor with length 500mm is bent into arc shape with a radius 260mm and helical curve. The experiment results verify the feasibility of separating method, which would provide effective parameters for precise 3D reconstruction model of shape sensing sensor.