Research On Running Monitoring System Based On Android

Life lies in exercise.With the development of society,people's demand for exercise and fitness is also more and more increasing.Using smartphones to monitor user's exercise and even guide their exercise have gradually become prevalent among people during running.Medical research shows that the stability of the rhythm between breath and step plays a crucial role during the running process.Therefore,it is of great practical significance to be able to implement a system for monitoring the running process.This thesis researches on the detection algorithm of step and breath based on the Android platform embedded sensors.In order to achieve the above-mentioned demand,this thesis researches on the detection algorithm of both step and breath,and achieves the following results:(1)Design and implement an effective step detection algorithm with the advantages of position-independent and non-interference features dependent on summarizing the existing step detection algorithm.Firstly of all,we use the Butterworth filter to eliminate high-frequency interference.Secondly,using the projection transformation method to solve the problem of smartphones of position-independent.Lastly,using a smooth filter to eliminate false peaks and valleys.Finally,we can achieve steps frequency.The algorithm is tested on the Android platform 7.0,and the results show that the algorithm has achieved good results in both energy consumption and accuracy.(2)Analyze the time domain and frequency domain characteristics of respiratory signals.By comparing the characteristics of short time energy,short-time zero crossing rate,Mel frequency cepstrum coefficient(MFCC),and linear predictive coding(LPC)parameters,finally,MFCC is selected as the parameter of model training.And then we use the support vector machine(SVM)to train the expiratory phase and inspiratory phase of MFCC parameters for the purpose of calculating the breath frequency.In order to improve the accuracy of the algorithm,this thesis researches on the error of classification points and noise interference,then propose a error of classification elimination based clustering algorithm and noise elimination based cycle algorithm,respectively.(3)We implement a system based on the Android platform for monitoring the state of exercise during running using the algorithm proposed in this thesis.Using the monitoring of early exercise rhythm and combined with respiratory frequency to choose a suitable rhythm song to adjust the pace of step in order to ensure the stability of the rhythm between breath and step.And achieve the effect ofthe efficiency of running and delaying the fatigue.