A method and system for lower limb motion recognition during human walking based on deep learning

Abstract

The invention discloses a deep learning-based lower limb motion recognition method and system for a human body walking process. According to the method, a wireless sensor is used for collecting leg myoelectric signals of a subject; filtration and standardization processing is carried out on the acquired myoelectric signals; time domain features and frequency domain features of the preprocessed myoelectric signals are extracted; the time and frequency domain features and original data are jointly used as inputs of a deep neural network model; the mixed input data is processed through the deep neural network model; the original data is processed by using a one-dimensional convolution neural network and a recurrent neural network; the frequency domain features are processed by utilizing a two-dimensional convolution neural network; the time domain features are processed by using the recurrent neural network; and finally a human body walking gait and joint angle recognition result is output after all the processed signals pass through a full-connection network. According to the method, the walking gaits and the joint angles of a human body can be quickly and accurately recognized, so that accurate wearer motion information can be provided for an exoskeleton robot.

Description

technical field [0001] The present invention relates to the technical field of computer pattern recognition, in particular to a method and system for recognizing lower limb movement during human walking based on deep learning. Background technique [0002] According to biological analysis, when the human body moves, the skeletal muscles contract and relax continuously with the movement of the limbs under the control of the cerebellum and brainstem. Carry out conduction, overlap in time and space, and form electromyography (EMG). By placing electrodes on the surface of the skin to obtain surface electromyography (Surface electromyography), the corresponding features can be extracted from the surface electromyography, so as to obtain the movement information of the lower limbs of the human body. Myoelectric signal is a complex non-stationary signal, which has strong ambiguity, is easily disturbed by the outside world, and is affected by multiple factors such as muscle fatigue...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a method and system based on deep learning for the recognition of lower limb movement in the process of human walking, so as to solve the problems that the traditional gait recognition method has low recognition accuracy for human subtle movements, poor real-time performance and cannot be continuously recognized.

Images