Eye Visual Recognition And Fatigue Monitoring Of Engineering Warch-keepers On Board Ship

According to statistics from the International Maritime Organization,seafarers’ fatigue duty has become a potential risk factor for ship navigation.More than 80%of sea traffic accidents have an indirect or direct relationship with fatigue duty.Although there are many different methods and techniques for detecting driver fatigue at present,it is mainly for the monitoring of the standing position of the cabin on duty or sitting,walking or sitting.This paper proposes to judge the fatigue state of the person in the cabin by checking the face state of the person on duty.Firstly,this paper uses the trained AdaBoost algorithm to combine the Haar-Like features to locate the face of the cabin shift personnel.On this basis,the improved AdaBoost algorithm is used to locate the human eye.Finally,the fatigue state of the cabin crew is monitored using the improved PERCLOS criteria.The main research content of this paper is divided into the following two aspects:(1)The AdaBoost algorithm combined with Haar-Like eigenvalues is used to identify the face of the cabin personnel.At the same time,based on the detection of the face,the AdaBoost algorithm is used to accurately identify the human eye.The research results show that compared with the traditional template matching algorithm,the algorithm can more accurately identify the face and human eye position.(2)It is proposed to use the improved PERCLOS fatigue criterion,that is,the method of closing the number of closed-eye frames per unit time to the total number of frames to monitor the fatigue state of the cabin shift personnel.The calculation results show that in the 15-second video recorded by the cabin duty personnel,one frame of the face image is acquired every second and the face person is recognized by the face,and the improved PERCLOS value is calculated by identifying the human eye state of the duty person.20%,finally determined to be awake.In this paper,a target recognition algorithm is proposed,which can quickly locate the face,and continue to quickly locate the eye position and recognize the eye state.It has the advantages of less computation,real-time,fast and so on.It can effectively monitor the fatigue state of duty personnel.It has important practical significance for improving the safety of maritime navigation.