Fatigue detection based on isophote curve

In order to improve the fatigue detection timeliness and accuracy, the optimization of isophote cure is used to locate the eyes, which reduces the amount of computation and improves the processing speed. First, according to facial features, the skin´s Gaussian model is established to differentiate face. The integral projection is used to locate the face. The area of eye is separated on the basis of three court five, and it uses the optimization of isophote cure to locate exactly eye. The gray value of the tiny area of eye is calculated and is used for the information of eye´s open and close, that is the PERCLOS value, and to determine whether one is fatigue or not. From the experiment, the system has a good timeliness and high detection accuracy.