Design Of Gait Recognition System Based On Plantar Pressure And Attitude Signal

With the rapid development of information technology,biometric identification technology is rising day by day.Through scientific and technological processing of the inherent physiological characteristics and behavioral characteristics of the body,people can realize identification of identity and identification of human movement behavior.Among them,gait recognition technology,as one of the representatives of biometric recognition technology,has the characteristics of remote recognition,non-invasive,low resolution requirements,etc.,and has been gradually applied in medical health,rehabilitation treatment,motion analysis,identity recognition and other fields.Based on the research background of daily monitoring of the elderly,this paper designs a portable wearable gait recognition system based on wireless body area network,aiming at the problems of single data signals collected in the existing recognition methods of gait recognition system and poor portability of the collected equipment.Based on the portable pressure and attitude sensor,the system can collect,transmit,process and analyze the foot pressure signal and attitude signal of human body in real time,which can realize the effective recognition of human gait.Firstly,aiming at the limitations of traditional gait recognition methods based on the surrounding environment,photoelectricity,electromyography and single sensor,this paper designs a multi-sensor,miniaturized and wearable gait information acquisition system.The acquisition module based on plantar pressure acquisition chip and nine-axis acceleration acquisition chip can realize synchronous acquisition of plantar pressure signal and attitude signal.Secondly,as for the limitation of power consumption and bandwidth of wireless volume domain network in information transmission,CC3200,which can be used for wifi wireless transmission,is selected as the main control chip in this project to meet the requirement of high-speed and high-precision signal transmission and communication.Thirdly,in view of the instability of collected gait signal features,this paper extracted the plantar pressure signal and attitude signal based on wavelet energy entropy feature and wavelet packet energy feature.Finally,in view of the limitations of the visual classification of gait feature graphs,this paper proposes a gait recognition method based on the extreme learning machine algorithm,which can effectively classify extracted features and recognize gait.Based on the human gait characteristics,six gait acquisition experiments were designed and carried out in the laboratory environment.The validity of the experimental data was confirmed through the calibration and testing of the pressure sensor,and the gait characteristics were analyzed.For the attitude signal,after the preprocessing based on the normalization of quaternion,the gait characteristics of each gait acceleration and the combined gait acceleration are analyzed,which lays a foundation for the subsequent research work of feature extraction.This topic with a gait recognition system based on wireless body domain network for the collection of data signals,proposed to use two kinds of the thinking of the energy feature of gait recognition,which is based on the wavelet energy entropy feature classification accuracy 86.62%,based on wavelet packet energy feature classification accuracy 94.37%,implements the effective of gait recognition,which laid a foundation for further research.