Bibliographic citations
Zamora, C., Vasquez, E. (2024). Desarrollo de un equipo electrónico vehicular orientado al monitoreo, detección y registro del estado de deterioro de pistas de tránsito urbano [Trabajo de suficiencia profesional, Universidad Peruana de Ciencias Aplicadas (UPC)]. http://hdl.handle.net/10757/683421
Zamora, C., Vasquez, E. Desarrollo de un equipo electrónico vehicular orientado al monitoreo, detección y registro del estado de deterioro de pistas de tránsito urbano [Trabajo de suficiencia profesional]. PE: Universidad Peruana de Ciencias Aplicadas (UPC); 2024. http://hdl.handle.net/10757/683421
@misc{renati/1039548,
title = "Desarrollo de un equipo electrónico vehicular orientado al monitoreo, detección y registro del estado de deterioro de pistas de tránsito urbano",
author = "Vasquez Vergara, Edwin Steef",
publisher = "Universidad Peruana de Ciencias Aplicadas (UPC)",
year = "2024"
}
The present work develops the implementation of a device based on embedded systems in a car aimed at monitoring urban roads and detecting potholes in order to obtain their location and severity level according to their size. Initially, a preliminary analysis was conducted on the inspection and monitoring method performed by the entities in charge of these functions. This analysis led to the conclusion that the method used is manual, subjective, and time-consuming. Therefore, it was proposed to automate the road monitoring process through the implementation of embedded systems based on Raspberry Pi 4B in the inspection car to detect, measure, and classify potholes simply by driving through the streets using algorithms based on ResNet50 and Deeplabview neural networks for pothole detection and diameter calculation, as well as an algorithm to obtain the coordinates of the detected pothole to locate its position and schedule a potential repair. The proposed solution involves recording the route using cameras, then storing the videos in a remote database. The video is downloaded onto a desktop computer for offline video processing to avoid computational overload. The software detects the potholes present in the video, calculates their size, and classifies them into severity levels based on diameter. As a result, 74.38% accuracy was obtained in pothole detection and 79.19% in size calculation.
This item is licensed under a Creative Commons License