The License Plate Recognition system for tracking stolen vehicles

For the License Plate Recognition system developed for tracking stolen vehicles, the system is designed to read the license plate of a car and compare it with the database. If the license plate number is found in the blacklist, the system will immediately send a warning notification. The system has been developed using Microsoft Visual Studio 2015 and VB.NET language, along with the EmguCV command for image processing and the K-Nearest Neighbor (K-NN) algorithm for number recognition on the license plate. During testing, the system has shown impressive performance. It can accurately read the license plate numbers when the license plate is positioned in the middle of the image and at a distance of 1 meter. The accuracy rate is 100% based on a sample of 30 images. Furthermore, the system is also capable of recognizing license plate numbers even when the license plate is tilted at an angle of 80 degrees or 100 degrees, and the distance is more than 1 meter. In this scenario, the accuracy rate is 50% based on a sample of 30 images. These test results demonstrate the system’s effectiveness in license plate recognition, particularly when the license plate is in the center of the image and within a certain distance range. Additionally, the system shows promising performance in recognizing license plates even in tilted positions, albeit with a slightly lower accuracy rate.