A gesture recognition method and device

Abstract

The invention discloses a gesture recognition method. The method comprises the following steps: a myoelectric signal of a target gesture object is collected through a plurality of channels, and the myoelectric signal is a one-dimensional time series of a motion unit in a muscle fiber when the muscle is contracted; the effective EMG signal is extracted according to the motion cycle of the preset gesture; the effective EMG signal is input into a linear predictive autoregressive AR model, feature vectors are extracted, the feature vectors extracted from a plurality of channels are fused, and thefeature vectors are combined into a multi-dimensional feature vector; the eigenvector is input into a preset standard centroid cluster K-means model, The clustering similarity between the eigenvectorand each centroid in the K-means model is analyzed, and the eigenvector is distributed into the centroid with the highest similarity to recognize the gestures. The invention also discloses a device based on the gesture recognition method.

Description

technical field [0001] The present invention relates to the technical field of gesture recognition, in particular to a gesture recognition method and device. Background technique [0002] In recent years, with the emergence and development of various human-computer interaction sensors in the field of computer application technology, in addition to the traditional keyboard and mouse input, more human-computer interaction methods have emerged. Among them, the human-computer interaction method based on gesture recognition is widely used in rehabilitation treatment, mechanical control and other fields because of its intuitive, concise, and convenient operation. The current gesture recognition methods are mainly divided into recognition methods based on computer vision, recognition methods based on pressure signals, recognition methods based on gyroscope sensors, and recognition methods based on EMG (electromyographic signals). [0003] Common recognition methods on the market, ...

AI Technical Summary

Problems solved by technology

Therefore, there is an urgent need for a new gesture recognition method to solve the problems that the existing gesture recognition methods are cumbersome to operate and the recognition accuracy of gestures is not high.

Images