| According to statistics, 13% of the operational errors in air traffic management are directly related to the controllers' fatigue; Among them, about 18% failures are related to the controllers' fatigue in ATC. Therefore, effectively monitoring the controllers' fatigue has great significance for ensuring safe operation of air traffic control and improving the ATC safety level.The main task of this project was designing a controller fatigue monitoring system, based on Open CV platform and studying key technology of facial recognition, to real-time monitor the controller's fatigue. Real-time images acquired by the camera controller, the obtained image signal is transferred to the PC for processing:1 ? Using the Adaboost algorithm, cascade classifier and ASM template matching algorithm for human face of rapid positioning;2?Firstly, choose the statistical learning based eye detection algorithm to determine the people's eyes and mouth initially rectangular area, followed by taking eye detection algorithm based on template matching ASM further refined to determine the human eye and mouth area.3?Use the local binarization and histogram equalization techniques to enhance and strengthen the country contrast to the human eye and mouth detection image into a black and white image. Then identify the state of the eyes and mouth. Comprehensively analyze and calculate to determine the controller's fatigue grade according PERCLOS value(Percentage of Eyelid Closure Over Time), average length of eyes closed, fatigue yawning frequency. ATC system alarms when ATCer in severe fatigue.Repeated experiments showing that fatigue monitoring system controller was reasonable designed, fully functional and accuracy especially for face recognition, eye position, eye and mouth state identification, and the rate of fatigue grades. The success rate was up to 85%. The system meets the monitoring needs. Moreover PERCLOS value and other data stored by systems makes difference on the controller fatigue risk management. |
PDF Full Text Request