Individual 3D Modeling Using Tilted Model Paradigm

Abstract

With the rapid development of 3D Geographic Information Systems (3D GIS) and smart city construction, oblique photogrammetry has been widely adopted for 3D reality modeling due to its efficiency and high precision. However, the triangular irregular network (TIN) models generated from oblique imagery typically form continuous surface models, making direct individualization and independent management challenging. This paper systematically examines the strengths and limitations of existing individualization methods and proposes a structured workflow for model individualization. By employing techniques such as model reconstruction and attribute association, the study achieves independent extraction and structured representation of targets (e.g., buildings, roads) within oblique models. Experimental results demonstrate that the derived models support subsequent spatial analysis and attribute management, providing enhanced data foundations for refined 3D geospatial applications.