Desarrollo de un Sistema-Prototipo para reconocimiento y detección de placas vehiculares usando visión e inteligencia artificial para el control de ingreso al parqueadero de la Gobernación de Sucre.

Abstract

Traditionally, parking lot entry controls are performed by security personnel, which can generate security problems due to human error or negligence. For this reason, we seek to design, build and validate a prototype system that identifies license plates through the use of vision and artificial intelligence. Artificial vision allows extracting information from images to take actions based on that data. Hardware wise, the system uses the Raspberry Pi 3b+ single board computer (SBC) in conjunction with a parallel-connected Pi Camera. On the software side, the image processing used for recognition is divided into three stages: plate detection, preprocessing and character recognition. The system results describe a plate detection rate of 100% and a character recognition rate of 58,33%.