Answers are grounded in this paper and cite their sources.
Urban mapping plays a key role in supporting effective planning and management. In this study, a practical workflow was developed to generate digital base maps from UAV-borne LiDAR point clouds using the Random Forest (RF) classification algorithm. The research was carried out in Italy City-2, Erbil, Iraq, where a dense LiDAR dataset captured by a Zenmuse L2 sensor was processed and classified into four main urban land-cover categories: buildings, roads, vegetation, and ground. The RF model was trained on radiometric and scanner-based attributes from a portion of the study area and then validated on an independent dataset. The results showed satisfactory performance, providing geometrically reliable and visually interpretable outcomes. Quantitative evaluation achieved an Overall Accuracy of 83%, a Macro-score of 0.81, and a Weighted-score of 0.83. These results demonstrate a highly efficient and cost-effective framework for municipal mapping in rapidly developing regions. The classified point cloud was finally converted into GIS-compatible vector layers, producing a detailed and thematically consistent urban base map suitable for municipal use. The proposed workflow offers a repeatable framework for urban planning, cadastral updates, and infrastructure evaluations—especially in rapidly developing areas where traditional spatial data sources are scarce.
No takes yet. Share an insight, caveat, or question.
Hama et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: