This paper has been proposed to reduce the probability of accidents of drivers who drive under the influence of drugs or alcohol. It proposes and effective model for drowsiness detection, and intelligent control or dimming of headlights when sensing a vehicle from the opposite direction. The proposed model shows an accurate algorithm to detect the Eye Map for eye detection, iris and pupil detection, and analysing the state of the eyes to make the drowsy decision. High precision techniques for location of iris, and pupil localisation helps in taking the important decision to ensure safety of the drivers. Intelligent Control of Headlight is implemented using an effective algorithm which uses SVM classifier for tuning the data to get correct detection of light patches from opposite direction. Finally, the paper shows the required connections for mounting a Raspberry Pi, and getting the model into effect, showing the various connections required for installing in a car.