PVBR-Recog: A YOLOv3-based Brazilian Automatic License Plate Recognition Tool

  • Pedro Ferreira Alves Pinto PUC-Rio
  • Antonio José G. Busson PUC-Rio
  • João P. Forny de Melo PUC-Rio
  • Sérgio Colcher PUC-Rio
  • Ruy Luiz Milidiú PUC-Rio

Resumo


Vehicle’s license plate detection and recognition is a task with several practical applications. It can be applied, for example, in the security segment, identifying stolen cars and controlling cars entry/exit in private areas. This work presents a Deep Learning based tool that uses the cascaded YOLOv3 to simultaneously detect and recognize vehicle plate. In experiments performed, our tool got a recall of 95% in plate detection and 96.2% accuracy in the recognition of the 7 characters of the license plate.
Publicado
29/10/2019
PINTO, Pedro Ferreira Alves; BUSSON, Antonio José G.; MELO, João P. Forny de; COLCHER, Sérgio; MILIDIÚ, Ruy Luiz. PVBR-Recog: A YOLOv3-based Brazilian Automatic License Plate Recognition Tool. In: WORKSHOP DE FERRAMENTAS E APLICAÇÕES - SIMPÓSIO BRASILEIRO DE SISTEMAS MULTIMÍDIA E WEB (WEBMEDIA), 1. , 2019, Florianópolis. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2019 . p. 121-124. ISSN 2596-1683. DOI: https://doi.org/10.5753/webmedia_estendido.2019.8149.