Reconocimiento automático de la placa de un vehículo de Ecuador

Abstract

This project, a system is developed that aims to recognize vehicle license plates in Ecuador, using computer vision techniques. As part of the preprocessing of the images, resizing and binarization have been carried out on the original images. For image segmentation, edge detection, connected components and a preliminary algorithm will be identified to identify the location of the letters and numbers on the board. Automatic recognition is carried out through two neural networks on the sections considered as "characters", taking into account the peculiarities of the Ecuadorian plates. The errors allowed bordering 3%.