SEMI-AUTOMATIC GENERATION OF AN LOD1 AND LOD2 3D CITY MODEL OF TANAUAN CITY, BATANGAS USING OPENSTREETMAP AND TAAL OPEN LIDAR DATA IN QGIS

3D city models have found purpose beyond simple visualization of space by serving as building blocks of digital twins and smart cities. These are useful to urban areas in the Philippines through diversified applications: urban planning, disaster mitigation, environmental monitoring, and policy making. This study explored the use of free and open-source software to generate an LOD1 and LOD2 3D city model of Tanauan City, Batangas using building footprints from OpenStreetMap and elevation models from Taal Open LiDAR data. The proposed approach consists of GIS-based methods including data pre-processing, building height extraction, roof identification, building reconstruction, and visualization. The study adopted methods from previous studies for building detection and from Zheng et al. (2017) for LOD2 building reconstruction. For LOD1, a decision tree classifier was devised to determine the appropriate height for building extrusion. For LOD2, a model-driven approach using multipatch surfaces was utilized for building reconstruction. The workflow was able to reconstruct 70.66% LOD1 building models and 55.87% LOD2 building models with 44.37% overall accuracy. The RMSE and MAE between the extracted heights from the workflow and from manual digitization has an accuracy lower than 1 m which was within the standards of CityGML. The processing time in test bench 1 and test bench 2 for LOD1 took 00:12:54.5 and 00:09:27.2 while LOD2 took 02:50:29.1 and 01:27:48.2, respectively. The results suggest that the efficient generation of LOD1 and LOD2 3D city models from open data can be achieved in the FOSS environment using less computationally intensive GIS-based algorithms.

QUALITY ASSESSMENT OF A NATIONWIDE DATA SET CONTAINING AUTOMATICALLY RECONSTRUCTED 3D BUILDING MODELS

Fully automated reconstruction of high-detail building models on a national scale is challenging. It raises a set of problems that are seldom found when processing smaller areas, single cities. Often there is no reference, ground truth available to evaluate the quality of the reconstructed models. Therefore, only relative quality metrics are computed, comparing the models to the source data sets. In the paper we present a set of relative quality metrics that we use for assessing the quality of 3D building models, that were reconstructed in a fully automated process, in Levels of Detail 1.2, 1.3, 2.2 for the whole of the Netherlands. The source data sets for the reconstruction are the Dutch Building and Address Register (BAG) and the National Height Model (AHN). The quality assessment is done by comparing the building models to these two data sources. The work presented in this paper lays the foundation for future research on the quality control and management of automated building reconstruction. Additionally, it serves as an important step in our ongoing effort for a fully automated building reconstruction method of high-detail, high-quality models.

Use of topographic LiDAR point clouds for building reconstruction

Demand for three-dimensional (3D) urban models keeps growing in various civil and military applications. Topographic LiDAR systems are capable of acquiring elevation data directly over terrain features. However, the task of creating a large-scale virtual environment still remains a time-consuming and manual work. In this thesis a method for 3D building reconstruction, consisting of building roof detection, roof outline extraction and regularization, and 3D building model generation, directly from LiDAR point clouds is developed. In the proposed approach, a new algorithm called Gaussian Markov Random Field (GMRF) and Markov Chain Monte Carlo (MCMC) is used to segment point clouds for building roof detection. The modified convex hull (MCH) algorithm is used for the extraction of roof outlines followed by the regularization of the extracted outlines using the modified hierarchical regularization algorithm. Finally, 3D building models are generated in an ArcGIS environment. The results obtained demonstrate the effectiveness and satisfactory accuracy of the developed method.

Use of topographic LiDAR point clouds for building reconstruction

Demand for three-dimensional (3D) urban models keeps growing in various civil and military applications. Topographic LiDAR systems are capable of acquiring elevation data directly over terrain features. However, the task of creating a large-scale virtual environment still remains a time-consuming and manual work. In this thesis a method for 3D building reconstruction, consisting of building roof detection, roof outline extraction and regularization, and 3D building model generation, directly from LiDAR point clouds is developed. In the proposed approach, a new algorithm called Gaussian Markov Random Field (GMRF) and Markov Chain Monte Carlo (MCMC) is used to segment point clouds for building roof detection. The modified convex hull (MCH) algorithm is used for the extraction of roof outlines followed by the regularization of the extracted outlines using the modified hierarchical regularization algorithm. Finally, 3D building models are generated in an ArcGIS environment. The results obtained demonstrate the effectiveness and satisfactory accuracy of the developed method.