Traffic parameters acquisition system using faster R-CNN deep learning based algorithm

Miguel Zinanyuca, Diego Arce

Producción científica: Capítulo del libro/informe/acta de congresoContribución a la conferenciarevisión exhaustiva

1 Cita (Scopus)

Resumen

Traffic parameters survey is important for proper control of traffic lights on the roads. Computer vision is one of the tools that offer greater advantages and lower cost compared to other alternatives. Particularly among the computer vision algorithms, the use of Deep Learning stands out against the traditional methods of image processing, due to the varying conditions of the environment. In the present paper, vehicle detection is performed by using a Deep Learning based algorithm, running the system trained under different environments for which the system was not trained. Later, an area of interest is defined in the image to be analyzed where, based on the detected vehicles, the necessary parameters of each of the routes of interest will be obtained. The parameters detection includes obtaining the queue lengths, estimating the average number of passengers in the region of interest and detecting the number of vehicles detected according to their type.
Idioma originalEspañol
Título de la publicación alojada2020 IEEE ANDESCON, ANDESCON 2020
EstadoPublicada - 13 oct. 2020
Publicado de forma externa

Citar esto