Irresponsible driving is one of the leading causes of mishaps & deaths in the world today. The present work aims at developing a preventive solution to this problem in the form of a driving aid for drunk drivers or people who are not in a fit state to drive. The said system proposes to enhance certain existing frameworks using image processing. Haar-wavelet classifiers & Speeded Up Robust Features (SURF) algorithm are used to analyze images of the drivers' eyes & patterns of movement to judge if they are eligible to drive. These complement the existing sensors, like the odor sensor, and help the system reach a confident decision. We plan to take this system further by creating collaborative MANETs among vehicles for preventing potential harm to other drivers as well.