Design Of Gait Information Recognition System Based On STM32

With the prosperity of society and the progress of science and technology,people pay more and more attention to their own health.At the same time,biometric recognition technology is also developing rapidly.Various characteristics of different parts of the body can directly or indirectly reflect the health of the body,among which gait is an essential and frequent behavior of almost everyone and has attracted much attention.Gait is a biometric feature of the human body,which is unique,easy to obtain,and difficult to be disguised.The derived gait recognition system not only plays an important role in the elderly monitoring,medical rehabilitation and other medical fields,but also can be used by the public security organs for identification.At the same time,it has also received attention from experts in the field of sports.At present,there are some problems in the methods selected in the common gait recognition system.For example,the number or type of sensors used is relatively small,so the data signal collected will be relatively small,resulting in no way to reflect the integrity of human motion.In addition,there are also problems such as the equipment used to collect data is not easy to wear,poor portability,low cost performance and unstable data transmission.In order to solve the above problems,this paper designs a wearable multisensor gait recognition system based on Wi-Fi transmission.The system selects a variety of sensors to collect the gait information of the human body,and transmits the gait information to the host computer,and then processes and analyzes the collected data to realize the successful recognition of human gait.Firstly,in terms of hardware,this design uses a film pressure sensor,a six-axis sensor and a myoelectric sensor to collect human gait information.The plantar pressure,the acceleration and angular velocity signals of the knee joint and the muscle electrical signals of the lower leg are collected respectively,so as to improve the integrity of gait information.The STM32F103ZET6 microcontroller is selected as the main control chip.In terms of transmission,the current very convenient and stable Wi-Fi transmission mode is used,and the ESP8266 module is selected to realize this function.Secondly,in terms of the software of the data acquisition part,one is the corresponding programming and integration of each sensor and Wi-Fi module in Keil software,and the other is the production of a multi-functional host computer in Lab VIEW software.The host computer can not only display the specific data collected by the sensor in real time and display it intuitively in the style of waveform diagram.The function of data saving and data playback is also added to improve the use value of the whole system.Finally,this design invited a number of volunteers to wear the gait recognition system to collect gait in different segments and continuous states respectively,and these data were saved and analyzed.For the piecewise collected discrete data,the support vector machine is used to classify it,and the average classification accuracy is 99.98%.For the continuous collected time series data,the long short-term memory network and the gated recurrent unit structure are used to identify it.Most of the recognition accuracy reaches 100.00%,and the convergence of the gated recurrent unit structure is better than that of the long short-term memory network,which verifies the feasibility of the whole gait recognition system.