Answers are grounded in this paper and cite their sources.
The distribution and availability of on-street parking in a city have important implications for transportation, economic, and environmental planning. However, few cities maintain comprehensive mapping of street parking beyond metered zones. Recent studies have proposed computational approaches to automate street parking prediction, but data availability remains a challenge. To address this challenge, this study introduces an integrated framework for mapping street parking using widely available Google Street View images. We develop a computer vision model to detect parking signs and a geometric algorithm to identify stationary vehicles from sequential street view images. A decision tree classifier combines sign and vehicle detection results to predict the presence of street parking at the road-segment level. Trained with 765 ground-truth labels combining official metered-zone records and manual annotation of street-view imagery, the classifier correctly identifies approximately 85% (103 out of 121) of the true cases in the validation set. The two detection approaches offer complementary information for predicting street parking beyond regulated zones, especially in residential areas. Our framework can support municipalities with limited databases in developing city-scale street parking maps.
No takes yet. Share an insight, caveat, or question.
Lee et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: