sEMG-Based Gesture Recognition Method for Coal Mine Inspection Manipulator Using Multistream CNN

Abstract

With the rapid development of intelligent mining technology, remote-operated underground inspection and rescue robot have been widely used. This article recognized the operator’s emergency gestures based on forearm surface electromyography (sEMG). First, a wireless six-channel sEMG acquisition device is built and a dataset named coal mine inspection manipulator gestures (CMMG) is acquired; then, the features of each channel signal are extracted to a 2-D graph by a continue wavelet transform (CWT) method. The multistream convolutional neural network (CNN) model is built to analyze the feature graphs so as to detect action segments. The comparative experiments showed that the method improved the accuracy and showed better performance on both the CMMG dataset and the public Ninapro DB1 dataset.