Sistema para reconocer la matrícula de vehículos mediante algoritmos de visión artificial

Abstract

Abstract: This graduation project addresses the growing need for automated vehicle license plate recognition due to increasing traffic, by proposing a system based on computer vision algorithms. This system has applications in traffic control, surveillance, road safety, and territory management. The methodology includes capturing vehicle images, detecting the license plate region, and recognizing characters using image analysis techniques and OCR. Computer vision algorithms, deep learning, and convolutional neural networks are employed to achieve accuracy. The outcome is a prototype that identifies vehicle license plates, with a specific focus on the city of Loja, Ecuador. The objectives include a literature review, identification of suitable datasets and algorithms, system development, and performance evaluation through usability metrics. This work contributes to enhancing road safety, traffic control, and intelligent territory management by automating real-time vehicle license plate recognition.