Robust Detection of Fatigue Parameters Based on Infrared Information

Driver fatigue is one of the major causes traffic accidents, and this need has increased amount driver detection systems in vehicles order to reduce human material losses. This work puts forward an approach based on capturing near-infrared videos from a camera mounted inside vehicle. Then, captured images using image-processing techniques eyes are detected. Next, features extracted eye several transforms finally, system detects if there or not SVM as classifier. Throughout recording, position tracked with low computational time analysed percentage eyelid closure. been developed two public datasets. Our experiments were able reach recognition rate up 96.87% our results showed that RBF kernel 99.66% accurate databases used for training. shows promising comparison state art deep learning approaches be implemented real conditions.