Identifying Spatio-Temporal Patterns of Traffic Congestion Using Data Obtained from Google Maps Service Traffic Image

Shahri, Matin; Mohaymany, Afshin Shariat · 2023 · Crossref

DOI: 10.52547/gisj.15.1.63

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of identifying spatio-temporal patterns of traffic congestion in Tehran, Iran, by leveraging big data derived from Google Maps traffic images. Traditional traffic data collection methods, such as physical sensors and license plate readers, are often costly, limited in spatial coverage, and prone to missing data. The authors argue that modern technologies, specifically digital traffic maps with historical data, offer a low-cost, comprehensive alternative for continuous spatio-temporal monitoring. The research aims to evaluate the performance of the transportation network at the city scale by analyzing congestion trends, identifying hotspots, and assessing temporal correlations. The methodology involved collecting traffic history images from Google Maps for a continuous one-month period (March 17 to April 17, 2019). Using a coding process, the researchers extracted 96 images per day (updated every 15 minutes) at a zoom level that covered all major arterial roads. Image processing techniques were employed to digitize color information, specifically isolating red pixels which indicate heavy or severe congestion. These pixel counts were aggregated to calculate a Traffic Congestion Index (CI) for each of Tehran’s 117 districts. The CI was computed as the percentage of red pixels relative to total pixels, averaged over hourly intervals to ensure meaningful temporal continuity. Spatial analysis utilized the Getis-Ord G* statistic to identify local spatial clusters (hotspots) of high congestion, distinguishing between working and non-working days. Temporal analysis employed the Kruskal-Wallis non-parametric test to evaluate the significance of differences in mean congestion values across four weekly time slices. Additionally, overlay analysis of traffic maps was used to identify persistent congestion clusters at a 90% confidence level. The results confirmed the presence of significant spatial and temporal correlations in traffic congestion values. The Getis-Ord analysis successfully identified specific districts with statistically significant high-congestion clusters, validating the existence of localized hotspots. The Kruskal-Wallis test rejected the null hypothesis of randomness, confirming temporal consistency and significant differences in congestion levels across different time periods and day types. Overlay analysis revealed distinct traffic clusters during morning and evening peaks, with variations observed between working and non-working days. The study demonstrated that Google Maps data, when processed through digital image analysis and spatial statistics, can reliably map congestion patterns at the district level. The significance of this research lies in its provision of a cost-effective, high-resolution method for urban traffic monitoring. The findings can inform the revision of traffic control zones and pollution control measures in Tehran. Furthermore, the identified spatio-temporal patterns provide a foundation for related studies on air pollution, toll pricing strategies, and the evolution of traffic bottlenecks over time. By validating the use of commercial digital map data for academic and planning purposes, the study offers a scalable approach for other metropolitan areas facing similar data acquisition challenges.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• traffic density