| With the improvement of people’s living standard and the transportation industry, the number of traffic accidents is more and more large. In frequent traffic accidents, driver’s fatigue is one main reason of the traffic accidents. Driving with fatigue caused by traffic accidents accounts for more than10%of the total number of traffic accidents.Therefore,how quickly and efficiently determine the state of driver’s fatigue has great significance on reducing traffic accidents.Firstly, the image preprocessing algorithm, target detection algorithm, fatigue judging algorithm is simulated.Secondly, the algorithm is transplanted through the comparative analysis of the optimal algorithm.The hardware implementation of system is based on ICETEK-DM6437-B development board of Beijing’s realtime innovation and technology limited liability company.The Software environment uses CCS (Code Composer studio)3.3development environment and uses its integrated real-time operating system DSP/BIOS,then teasily do the real-time scheduling and real-time debugging.This article focuses on the image preprocessing, and the target detection of human faces, eyes and the judgment algorithm on the basis of it,as well as the realization of it in the DSP.Image preprocessing module adopts the median filter and the image histogram equalization. Targeting detection module is based on skin color in the algorithm,and geometry character method of target detection, fatigue judgment in the module is based on the PERCLOS algorithm to realize the judgments of the fatigue of the driver. The whole system realize the basic goal of the fatigue test function,and the system is feasible for the person with glasses and without glasses for target judge, and the system has high accuracy.At the same time, the system is not only applied to the pilot fatigue detection, but also applied to some monitoring room, large control room of the operator of the occasion to detect fatigue, to reduce great loss caused by fatigue. |
PDF Full Text Request