Published Paper Details:

of the most critical problem overcome in India is death caused by the road accidents. Almost 80% of accidents are happen due to driver distraction. We attempt to develop robust system for detecting Driver distraction. There are some methods to detect it but that consumes more time. It is important to early detect the distraction, inform to the driver about the distraction. So, computer-based application needs to be developed to detect this distraction as early as possible and minimize the risk of accidents. The aim is to develop a simple and capable method to detect the distraction. There are some proposed methods that contain following stages, preprocessing, feature extraction and classification. To increase the accuracy of the result we use two layer neural network. ELM, softmax are used for feature extraction and SVM-ELM and method is used for classification. The accuracy of proposed method is 96.7% which shows its reliability.