Abstract:
Vehicle Identification is being done by Vehicle Number Plate Recognition (VNPR) systems in applications such as traffic monitoring, toll collection, and security surveillance. However, many existing VNPR systems are intended for large-scale environments such as toll gates and city-wide monitoring infrastructures. This paper presents a VNPR system which is simple and cost-effective majorly designed for smaller regions such as public parks, institutional campuses, and restricted residential buildings.
The system automatically captures vehicle images either from stored input images or through a live camera interface. Image processing techniques is used to analyse vehicle images to detect the region of number plate and extract the characters through a technology called Optical Character Recognition (OCR). In addition, a Convolutional Neural Network (CNN) based deep learning approach is used to enhance the performance of character recognition.
The recognized vehicle number can further be used for access control, vehicle monitoring, and maintaining entry records within the premises. The proposed system therefore provides an efficient automated solution for vehicle identification in small and controlled environments while reducing manual effort and improving security .
Keywords: