Detection On Fatigue State Based On Fusion Of Multi-machine Vision Technique

With living rhythm more and more quickly,people have less and less time to rest each day,which makes it difficult to be able to have adequate energy into the daily work and learning.In this circumstance,ability to work will be deskilled,even more it may cause great economic losses and casualties in a particular industry.Especially in field of aerospace,long-distance transport,large servers regulatory,once the operator work in mental fatigue state,it will make the entire system in a very dangerous situation.Recently,most of mental fatigue detection system have many weaknesses,which are unacceptable in these industry.Therefore,a new fatigue detection system is urgently needed.A new method of fatigue state detection is presented in this paper,which fuse multi-technique based on machine vision.Firstly,it presents a new technique of heart rate measurement based on machine vision.In order to improve the accuracy of the heart rate measurement,two new method are proposed: average in frequency and a cubic spline interpolation reconstruction method based on actual schedules.What's more,experiments of simulation and real heart rate measurement show that the algorithm is sufficiently effective to enhance both the amplitude and frequency precisions.Secondly,use the cam-based heart rate measurement technique in heart rate variability that can be applied in fatigue detection,and experiments of fatigue detection proof the properties of the method.Therefore,time domain and frequency domain of heart rate variability is concerned to the fatigue state detection.Finally,the technique fused cam-based heart rate measurement and P80 measurement is tested by experiments in different objects and it shows that average in time domain and high-frequency component in frequency domain of the heart rate variability are related to fatigue state.And the P80 is more relevant to the deep-fatigue.A new model fused with multi-characteristic of fatigue is presented in this paper,which can be applied to different industries by adjusting the parameters.As a result,the fatigue state detection system can be used as a monitor of fatigue state and early warning system of fatigue,which has advantages of simple operation,low cost,non-invasive,etc.and has great value in practical application.