Pedestrian crossing detection method in urban traffic using deep learning

Pedestrian crossings, along traffic lights, are an important element of pedestrian safety in urban traffic and suburban roads. Similarly, traffic information is crucial to improving the safety of these vulnerable road users. This paper presents a method for automatic pedestrian crossing detection in car camera images. The method can also be applied using UAV or satellite imagery. A deep learning methodology using a convolutional neural network (YOLOv4) was used to process real-world images. Images obtained from video frames of pedestrian crossings observed at different distances were used to train and test the deep learning model. For test images, which were not part of the training sequence, an accuracy of 77% was achieved for a detection threshold greater than 0.5 and an accuracy of 100% for a threshold greater than 0.01. The developed model can be used to detect pedestrian crossings in various vehicle control applications and in autonomous vehicles. The located objects in the images can also be analyzed for their quality. Crosswalks may be partially dirty or partially worn and may require renovation.

Authors

Institutions

Publication Details

Journal
Journal of Civil Engineering and Transport
Published
2026-11-16
DOI
https://doi.org/10.24136/tren.2026.011
Primary Topic
Advanced Neural Network Applications
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Pedestrian crossing detection method in urban traffic using deep learning

Zbigniew Pamuła
1 links
Journal of Civil Engineering and Transport
Advanced Neural Network Applications
article

Pedestrian crossing detection method in urban traffic using deep learning

Zbigniew Pamuła
article en

Abstract

Pedestrian crossings, along traffic lights, are an important element of pedestrian safety in urban traffic and suburban roads. Similarly, traffic information is crucial to improving the safety of these vulnerable road users. This paper presents a method for automatic pedestrian crossing detection in car camera images. The method can also be applied using UAV or satellite imagery. A deep learning methodology using a convolutional neural network (YOLOv4) was used to process real-world images. Images obtained from video frames of pedestrian crossings observed at different distances were used to train and test the deep learning model. For test images, which were not part of the training sequence, an accuracy of 77% was achieved for a detection threshold greater than 0.5 and an accuracy of 100% for a threshold greater than 0.01. The developed model can be used to detect pedestrian crossings in various vehicle control applications and in autonomous vehicles. The located objects in the images can also be analyzed for their quality. Crosswalks may be partially dirty or partially worn and may require renovation.

Journal of Civil Engineering and TransportVol. 8(3)
Silesian University of Technology (PL)
Sustainable cities and communities, Industry, innovation and infrastructure
Advanced Neural Network Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon