Gait Recognition Based On Static Foot Pressure

The basic principle of gait recognition based on static characteristic of plantar pressure is that different people’s standing foot pressure distribution under the condition of static is unique and stabile during a given period of time. Achieve the purpose of classification through analysis and extract different individual sole static pressure data, find out high degree of differentiation to achieve purpose. The basic principle of gait recognition based on static characteristic of plantar pressure have the advantages of long distance, and difficult to imitate and targeted crack, is a new thought of biometrics.This paper discusses the use of gait tactile characteristics for identification of medical and behavioral science theory, and discussed how to use gait tactile feature to complete identity recognition from the aspects of software and hardware, the full text main research work is as follows:1) A pressure measurement system based on pcap01 is developed, by measuring the plantar pressure deformation caused by the formation of conductive cloth capacitance change to measure plantar pressure information, measurement platform using stc12 single chip microcomputer to control, use a wireless blue tooth serial measurement platform to obtain the complete gait tactile information upload first place machine. Establish the initial gait pressure characteristics of the data set.2) Study on how to effectively remove the noise, select 0 resistance in hardware shielded wire, double power supply mode of the noise reduction processing of the circuit, to deal with the noise in the system, pressure data by wavelet transform threshold, unified by coordinate rotation invariant principle, to obtain the pre processed data.3) Methods of feature extraction and classification is discussed, the extraction of11 representative feature points as the basis, find out the distance and angle between these points, a plantar pressure characteristics of the first frame, using principal component analysis method processing the original features, and then use the genetic algorithm into the step of feature selection, and finally through the KNN data setclassification.Experimental results showed that the design of plantar pressure acquisition system can work normally, can have to static human plantar pressure parameters accurately capture function, and can be more quick and accurate data to upload the first machine, in order to ultimately achieve the purpose of identification, KNN classification recognition rate as high as 0.85 finally, verify the effectiveness of the algorithm.