Gait recognition and intention perception method

Abstract

The invention discloses a gait recognition and intention perception method, and the method comprises the steps: employing a data collection system which is used for collecting the plantar pressure, ankle joint acceleration and angular velocity information; the method comprises the steps of establishing an actual measurement database by utilizing actual data information acquired by a data acquisition system; performing human body model motion simulation by utilizing AnyBody software to obtain corresponding relations among plantar pressure, ankle joint acceleration, angular velocity informationand corresponding human body postures, and establishing a simulation database; estimating a mapping relation between an actual measurement database and a simulation database through feature extractionand feature matching, gait recognition and intention sensing are performed by using a convolutional neural network algorithm, and the current posture of a human body can be obtained under the condition that the posture of the human body at a place is not clearly reflected, so that the purposes of gait recognition and human body intention sensing are achieved. The problems that in the prior art, the pressure collection range is not comprehensive, the feature is single, the human body posture cannot be predicted, and the real-time recognition accuracy and speed are low are solved.

Description

technical field [0001] The invention relates to the technical field of biometric identification, in particular to a method for gait identification and intention perception. Background technique [0002] The control strategy of power-assisted exoskeleton robots is a hot research topic at home and abroad. Accurately predicting the state of human walking is the premise and basis for the control system to have predictability and rapid response capabilities. Gait information such as pressure, joint acceleration, and joint angular velocity are used to identify and judge the walking state. At present, the methods of collecting plantar pressure mainly include pressure plates, test benches and thin-film pressure sensors, but the pressure plates and test benches are not convenient to move, and real-time dynamic walking data cannot be obtained. The pressure range obtained by installing thin-film pressure sensors on special parts of the soles of the feet Not comprehensive, may lose key...

AI Technical Summary

Problems solved by technology

The method of gait recognition and intention perception establishes the mapping relationship between the measured database and the simulation database, and uses AnyBody software to perform human model motion simulation to obtain the corresponding relationship between plantar pressure, ankle joint acceleration, angular velocity information and corresponding human posture. Finally, the convolutional neural network algorithm CNN is used for gait recognition and intention perception. When the posture of the human body is not clear, the current posture of the human body can be obtained, so as to achieve the purpose of gait recognition and perception of human body intentions, and overcome the current situation. Existing technologies have the problems of incomplete pressure collection range, single feature, inability to predict human posture, and low accuracy and rate of real-time recognition

Images