Font Size: a A A

A Study On Driver Fatigue Detection System Based On DM642

Posted on:2014-04-13Degree:MasterType:Thesis
Country:ChinaCandidate:L WangFull Text:PDF
GTID:2252330422965770Subject:Electronics and Communications Engineering
Abstract/Summary:PDF Full Text Request
With the rapid rising number of vehicles, the rate of traffic accidents is climbing up,doing great harm on people’s life, property and health. Driving fatigue is one of factors tolead to the traffic accidents. Thus, the detection and control of traffic accidents caused bydriving fatigue is not only valuable in academy, but also important in social sense.Therefore, designing an efficient, real-time detection system becomes a key technology toreduce the rate of traffic accidents.With the help of the high speed and real time in DSP6000series, here we designed afacial fatigue detection system based on machine vision in the core processor digital signalprocessor (DM642), made by TI company, and established the system of hardware andsoftware for the system. The main contents are as below:(1) We proved the feasibility to design the driver fatigue detection system based onDM642. Using the principle of infrared light source imaging, we adopted the active modeof infrared imaging. We proposed a feasible processing algorithm, and then designedmodules of hardwardware circuits to make data flow.(2) We built the processing platforms of hardware and software based on DM642andthen complished them on DM642simulation system. The hardware system includes someparts, such as acquisition circuit modules, DSP blocks, power module, and so on. In thesoftware design, we designed the DSP programs, embedding both the algorithm to extractthe facial feature vector and the algorithm of support vector machine (SVM) to detect thefatigue.(3) The system has been debugged and analyzed, it gives the results of theexperimental tests and could operate stablely and reliablely. And significantly improve thereal-time processing capabilities and the efficiency of detection in the driver fatiguedetection system, the processing speed of an image could be15ms. Both the applications ofDSP/BIOS operation system and the technology of multiful parallel pipe-linescommunication can not only grantee the reliability and stability in the system operating, butalso lower the developing cost.
Keywords/Search Tags:Real-time, embedded system, Driver fatigue, DM642, Feature vector, extraction SVM
PDF Full Text Request
Related items