Research On Dynamic Gesture Recognition Technology Based On Planar Capacitive Sensing Array

In recent years,the Rapid Development of the Internet Industry,human-computer interaction technology is becoming more and more common,as a more in line with human habits of human-computer interaction technology,gesture recognition technology has been more and more valued by scholars from all walks of life.At present,the more popular gesture recognition technology is greatly affected by environmental factors such as light and weather,while the recognition technology based on the plane capacitive sensor has strong robustness,and as a non-contact sensor,it is more suitable for dynamic gesture recognition.In this thesis,the design of planar capacitive sensing arrays and signal acquisition and processing are studied.The planar capacitive sensing array dynamic gesture recognition system designed in this thesis has high stability and recognition rate.First of all,starting from the principle,the characteristics of the electromagnetic field and the dielectric are analyzed theoretically,the induction mechanism of the planar capacitive sensor is theoretically analyzed,and the factors affecting the sensing performance of the sensor are given.Based on the above theoretical research,the four performance indicators of the planar capacitive sensor are introduced in detail,and the influence of the major parameters of the planar capacitive sensor on the sensor performance is obtained through the COMSOL simulation software,and the capacitance change of the sensor for different hand positions is simulated,the feasibility of the position recognition of the opponent position of the planar capacitive sensor array is determined,and the crosstalk problem between adjacent sensors is also studied,and it is decided to add electrode shielding between adjacent sensors.Finally,the parameters of the capacitive sensor were determined to be a spiral structure with a radius of13 cm and an electrode width and electrode spacing of 1 cm.The array structure is designed with a regular hexagonal array with a total of 7 sensing units.The current small capacitance detection methods are analyzed and compared,and finally the AC lock-in amplification capacitance measurement circuit is used,and the scheme design of gesture signal acquisition hardware circuit is carried out according to the selected capacitance measurement circuit,which mainly includes FPGA main control circuit,excitation signal source design,digital detector module design,signal acquisition and Ethernet signal transmission design.The collected position data is affected by ambient white noise and crosstalk of adjacent sensors,and the resulting position data needs to be filtered by Kalman to make the filtered data smoother.Finally,by using the hidden Markov model(HMM),the parameter matrix of the HMM model is trained in advance by different gestures,and the dynamic gesture is matched by calculating the probability of the current hand position sequence for different HMM models during recognition,and the training and probability calculation of the HMM algorithm are realized through MATLAB.Finally,the feasibility of the gesture recognition system was verified by experiments,and the recognition rate of simple gestures could reach 100%,and the average recognition rate was higher than 95%.