Surveillance System using Moving Vehicle Number Plate Recognition

A growing popularity of self-owned vehicles in the modern world has called for an even increased demand for better enforcement of traffic rules, often at the traffic signals and important crosses of crowded livelihoods. Due to the increase in the number of self-owned vehicles there is a huge demand of such systems for the surveillance of the vehicles either it can be or traffic purpose or it can be used as surveillance system near a sensitive zones like military base, parliament and other key headquarters. Proposed road safety personnel are in need of helping hand from the software world to develop tools that can recognize, with very high accuracy, the number plates of the vehicles bring up the details related to their registration number. The above state problem can be resolved using an Automatic Vehicle Number Plate Recognition system (ANPR). There are several research goings on in this field to make ANPR systems more accurate and liable. Although the moto of this study is going to demonstrate simple ANPR system which can detect number plates efficiently using object detection methodology. The given problem is to design a model which would detect a vehicle number plate automatically and in addition to that extract the text and number from the detected image with the help of Optical Character Recognition (OCR). So, it's a combination of both object detection and OCR. YOLO object detection model is used with Py-tesseract engine for text recognition. ANPR system using YOLO detection model has shown significant accuracy in license plate detection and can be used for ANPR systems. Proposed system is detecting license plate with an accuracy of 97.0 %. The other parameters like precision, map, and recall values are also satisfiable.