The Detection Of Facial Fatigue By Infrared Imaging System Based On DM642

At present ,most of detection methods are limited to the static images,which greatly reduce the dfficiency and the temporal nature and take much more time of detection and judgement.not available in real-time system.Furthermore ,current video devices of image acquisition are used by visible light camera,which strictly limited by light strength around the surronding.When the background light is weak,or largely changed,the video images will not suitable to its pro-process.Here,some technology of image processing such as background subtraction and the infrared camera are employed to capture the moving target and reduce the reliability on the background.in order to improve the efficiency of the imaging system. The project is made up two sections:First section:the modules to detect the fatigue are developed with the help of image detection based on MATLAB/Simulink.Since the infrared camera is not highly relay on the background light so that there is an obvious edge to be able to be extracted.which is suitable for the pro-process.The method of the background subtraction is employed to extract the moving target after extracting the background image.Then,the motion images are binaried to calculate the largest connected region of the whole image in order to find the face area.Then the integral projection algorithm is utilized to locate of the human eyes.The PERCLOS principle is to judge the fatigue.At last ,all the algorithms and simulation modules are built in MATLAB/Simulink.Second section:C codes are produced and run on DM642.All simulation modules are transferred into C codes with the help of Real-Time Workshop in MATLAB/Simulink which is needed by DM642.Meanwhile,they are debugged and compiled in Code Compile Studio (referred to as CCS).Finally ,download and run the codes in DM642 to finish the verification of the whole system.