Abstract
In today’s scenario, automobiles have been a huge part of our day-to-day life and also have a huge impact on the nation's economy. But with the increase in the number of automobiles, there is a rise in accidents too. The use of a high beam in front of a coming vehicle, creates a glare on the eyes of the driver which makes him/her partially blind for a few microseconds, which is enough for an accident to take place. This paper proposes an Automatic High to low beam adjuster that adjusts the beam according to the presence of the car in front of it. We are using deep learning and a masking approach independently for detection of a vehicle in front of the primary vehicle whose beam we would be adjusting. The AI model would then be connected with an Arduino by which we would be converting digital signals to electrical signals. The scope of the project is to be able to detect headlights of oncoming vehicles and adjust the beam of the vehicle (high to low and vice versa) without the driver’s intervention as it would be of great help to the people driving at night, aged people and people with vision problems like cataracts etc. It could bring a whole new dimension of traffic control and road safety by detecting the headlights of the vehicles using dynamic footage recorded by sensors in real time. This paper will also be of crucial importance in reducing the number of accidents therefore preventing mishaps, saving lives and preventing financial losses. The future scopes could be integrating Raspberry Pi and using it with an Arduino and a camera to develop and test the fully functional product which can then be installed in vehicles to control accidents. The camera that would be used to detect oncoming vehicles could also detect emergency factors like sudden accident-like situations and could take preventive measures to reduce the impact or the probability of an accident.