Automatic Parking System using Vehicle License Plate Detection

With phenomenal increase in the number of vehicles, vehicular systems and parking systems are a major challenge faced by urban cities. Parking systems currently rely on manual labour for noting down the registration numbers of vehicles entering the parking system. Our project aims at developing a parking system what would detect a vehicle while entering in the parking lot, and also would automatically recognize its registration number at day and night time. Parking lots are generally closed structures so there is always low light and at night time there are extremely low light conditions. Any camera would fail to provide noise free image at this condition. So there is need for low light enhancement. This paper discusses some of the existing low image enhancement algorithms. From the enhanced image, computation is done in detecting the vehicle, classifying it into4 wheelers or 2 wheeler vehicles and also extracting the registration number from the detected license plate. This paper deals with several existing architectures and models which perform ALPR on benchmark dataset.