RT Journal Article
T1 Using Sensor Fusion and Machine Learning to Distinguish Pedestrians in Artificial Intelligence-Enhanced Crosswalks
A1 Lozano Domínguez, José Manuel
A1 Redondo González, Manuel Joaquín
A1 Dávila Martín, José Miguel
A1 Mateo Sanguino, Tomás Jesús
AB Pedestrian safety is a major concern in urban areas, and crosswalks are one of the most criticallocations where accidents can occur. This research introduces an intelligent crosswalk, employingsensor fusion and machine learning techniques to distinguish the presence of pedestrians and drivers.Upon detecting a pedestrian, the system proactively activates a warning light signal. This approachaims to quickly alert nearby people and mitigate potential dangers, thereby strengthening pedestriansafety. The system integrates data from radio detection and ranging sensors and a magnetic fieldsensor, using a hierarchical classifier. The One-Class support vector machine algorithm is used toclassify objects in the radio detection and ranging data, while fuzzy logic is used to filter out targetsfrom the magnetic field sensor. Additionally, this work presents a novel method for the manufactureof the road signaling system, using mixtures of resins, aggregates, and reinforcing fibers that arecold-injected into an aluminum mold. The mechanical, optical, and electrical characteristics weresubjected to standardized tests, validating its autonomous operation in real-world conditions. Theresults revealed the system’s effectiveness in detecting pedestrians with a 99.11% accuracy and a 0.0%false-positive rate, marking a substantial improvement over the previous fuzzy logic-based systemwith an 81.33% accuracy. Attitude testing revealed a significant 33.33% reduction in pedestrianerratic behavior and a substantial decrease in driver speed (32.83% during the day and 70.6% duringthe night) compared to conventional crossings. Consequently, this comprehensive work offers aunique solution to pedestrian safety at crosswalks by showcasing the potential of machine learningtechniques, particularly the One-Class support vector machine algorithm, in advancing road safetythrough precise and reliable pattern recognition.
PB MDPI
SN 2079-9292 (electrónico)
YR 2023
FD 2023-11-21
LK https://hdl.handle.net/10272/22705
UL https://hdl.handle.net/10272/22705
LA eng
NO Lozano Domínguez, J. M., Redondo González, M. J., Davila Martin, J. M., & Mateo Sanguino, T. de J. (2023). Using Sensor Fusion and Machine Learning to Distinguish Pedestrians in Artificial Intelligence-Enhanced Crosswalks. In Electronics (Vol. 12, Issue 23, p. 4718). MDPI AG. https://doi.org/10.3390/electronics12234718
NO This work was financed by the General Secretariat of Universities, Research and Technologywithin the scope of the Andalusian Research, Development and Innovation Plan (PAIDI 2020) andthe European Regional Development Fund.
DS Repositorio Institucional de la Universidad de Huelva
RD 14 jul 2026