Citas bibligráficas
Ruiz, S., (2023). Sistema de estimación automática de ocupación de carriles y trayectorias seguidas en avenidas [Tesis, Universidad de Ingeniería y Tecnología]. https://hdl.handle.net/20.500.12815/312
Ruiz, S., Sistema de estimación automática de ocupación de carriles y trayectorias seguidas en avenidas [Tesis]. PE: Universidad de Ingeniería y Tecnología; 2023. https://hdl.handle.net/20.500.12815/312
@misc{renati/230564,
title = "Sistema de estimación automática de ocupación de carriles y trayectorias seguidas en avenidas",
author = "Ruiz de Castilla Mendoza, Sebastián",
publisher = "Universidad de Ingeniería y Tecnología",
year = "2023"
}
In topics of transport management to alleviate traffic congestion and achieve fluid traffic, it is necessary to know the volume of vehicular traffic on the corresponding roads. In Peru and other developing countries, traffic volume is obtained mainly through manual measurement methods of vehicular capacity, characterized by their imprecision and susceptibility to human errors. This work presents an automatic estimation system of lane occupancy and trajectories followed by vehicles on avenues. This system consists of a camera and a computer that performs the video processing locally, without cloud computing. The system computes the area occupied by the vehicles and estimates the traffic volume in passenger car equivalents (PCE), an equivalence metric used in contexts of heterogeneous traffic or with different types of vehicle, without detecting each one of them individually. Thus, the system has a low computational consumption compared to current methods. This system can be used even in avenues that do not have a camera installed. In tests carried out with previously recorded videos, the system obtained a root-mean-square error of 1.42 % without training, equivalent to omitting 0.17 PCE (passenger cars) in 1 minute of measurement, compared to 6 PCE omitted in the same time with manual measurement. Then, after training of the system, the error was reduced to 1.20 %. In the tests carried out in the field in real-time, the error was 2.38 %.
Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons