Regularizing Vehicle Legalities through Machine Learning Techniques

This project aims to check the vehicle-related legalities i.e., insurance, registration related matters-license, grant certificate of fitness for Goods Carrying Vehicle (GCV), Passenger Carrying Vehicle (PCV), private vehicles and two wheeler's. This work would be done by petrol bunk workers by screening the registration number on the plate. A camera is used to capture the registration plate from which the number is identified using an image processing system that can deal with low illuminated, cross angled, nonstandard font number plates by employing techniques such as, morphological transformation, Gaussian smoothing, Gaussian thresholding in the pre-processing stage and sent to the respective servers. This license plate recognition is done through the Convolutional Neural Network (CNN) model which is further connected to the servers which checks for the legalities by redirecting to the insurance servers and RTO servers respectively. After the same, the vehicle holders are warned and fined accordingly based on their vehicle legalities perfection. This fine receipt is redirected to the respective servers for future use. This is one of the necessary systems designed to detect the vehicle number plate to get information about the vehicle legalities. This system illustrates a design and development for a new effective way for regularizing vehicle-related legalities using machine learning techniques. This system is implemented at the petrol bunk.