This research paper includes about the driver’s safety. It has three different models. They are Drowsiness detection system, Lane detection system and Vehicle maintenance system. Drowsiness is one of the major reasons for road accidents. To overcome this, prediction of drowsiness is developed. In fact, the drowsiness affects the performance, capability, physiological indices. These parameters provide information about driver’s state. First, the camera captures the driver’s face and then drowsy state is recognized by the machine learning algorithms. As a result, an alert can be initiated. Second, while lane changing one has to be very conscious, lane changing is one of the major cause of accidents. Computer vision and machine learning will be used to understand the change of lane, any changes in the path will be notified to the driver and an alert will be initiated. In the paper, we are using canny edge detection, gaussian smoothing and hogs transform which will increase the efficiency of the system as a result we can avoid the road accidents and ensure the safety. Third model in this paper is vehicle maintenance system. Here, based on the previous maintenance records, date and time of last service, the system will predict the next date of maintenance. We are classifying the model using Sklearn package and support vector machine algorithm. The generated model can predict and perform at least an accuracy of 90% and can be improvised by using the large datasets.

Machine learning drowsiness detection algorithms canny edge detection hogs transform gaussian smoothing computer vision classification Sklearn tools and support vector machine (SVM).