Sistema para reconocer señales reglamentarias de tránsito mediante algoritmos de visión artificial

Abstract

Abstract: The present work focuses on the development of a signal recognition system traffic regulations using artificial vision algorithms. The main objective of system is to improve road safety by allowing vehicles to automatically detect and recognize traffic signs autonomously, which will allow them to make informed decisions and avoid possible accidents. The data set used in the development of the system consists of 1086 photographs of traffic signs, acquired from both online and offline sources. photographs captured by the author to ensure a varied representation of situations. The system is trained with traffic signs to improve its recognition capacity in different conditions depending on their visibility. The model is validated with a data set test and is implemented in an artificial vision system that processes images in time real and recognizes traffic signs. The work includes a bibliographic review and analysis of related previous works, as well as tests and evaluations of the developed system. He work concludes that the vision-based traffic sign recognition system artificial is an innovative and effective solution to improve road safety in our cities.