Low-cost sensor solutions provide a great opportunity for democratization of mapping among different communities including those working in digital twin areas. These sensors provide relatively dense point clouds with limited accuracy especially in the absence of GNSS. In this study, we provide two solutions for improving the accuracy of the final point clouds produced by iPhone-based LiDAR and images. One solution utilizes length observations as constraints in the network. Our tests showed that this solution improves the relative accuracy of the point cloud for applications such as culvert mapping where absolute accuracy is not of high necessity. In the other case, we studied the effect of using ultrasonic position systems to aid with the absolute accuracy improvement in the network. The results of this study are still under investigation and will be published in the full paper.

Building accurate indoor digital twins is essential for smart-building services such as asset tracking, space planning, and AR/VR navigation. Yet a single LiDAR sensor cannot supply both millimetre-level accuracy and complete coverage: fixed terrestrial laser scanners (TLSs) leave occlusion holes, whereas handheld mobile laser scanners (HMLSs) suffer from lower geometric stability and colour drift. We propose an adaptive voxel-based fusion pipeline that combines a Faro Focus X330 TLS with a CHCNAV RS10 HMLS to overcome these limits. First, the handheld point cloud is rigidly registered to the TLS reference using Iterative Closest Point. TLS voxels with sparse points are flagged as holes; for each hole we admit only handheld points whose point-to-plane distance and normal deviation fall below strict thresholds, ensuring geometric consistency. Next, we correct colour bias by learning a global linear RGB mapping from overlapping scans and refining it locally with weighted regression. Finally, we blend colours across the TLS–handheld boundary to remove visible seams. Experiments on classroom scenes from a smart-campus testbed show that our method recovers 85.7% of missing surfaces, lowers the global point-to-plane RMSE by 14.8%, and improves mean colour difference by 22.2%. The resulting high-fidelity, colour-consistent indoor models give facility managers and planners reliable data for maintenance scheduling, occupancy analysis, and long-term space optimisation.

This paper presents the 3DTiler tool, which supports the manual georeferencing of 3D building models in .glTF or .glb format. Through a browser based interface, users can upload their 3D models and interactively position them on a 3D globe by setting geographic coordinates, elevation, and rotation. The tool then generates a 3D Tile Tileset, which can be used to integrate the models into OGC 3D Tiles-based platforms. The paper outlines how 3DTiler is designed as a lightweight, client-side WebGIS application that facilitates accurate model placement without requiring a backend server or cloud processing. It is part of the University of Cologne’s broader efforts in projects like CampusGIS2 and Virtual Campus to build immersive 3D and XR applications. The tool will be made open source and hosted on university infrastructure.

It is desirable that a 3D city model at level of detail 2 (LOD2) has texture images mapped on building surfaces. Texture images mapped on building surfaces are utilized not only for providing a realistic appearance but also for enabling some analyses, such as investigation of stories of a building and detection of doors and windows on a building surface. We conducted an investigation on the resolution of texture images produced from aerial nadir and oblique images. First, we set up the required resolutions for the investigation of stories and the detection of doors and windows by using simulated images of various resolutions. Next we obtained the general formula to calculate the resolution of a texture image mapped on a vertical surface. Then we conducted a numerical analysis on a texture image mapped on a vertical surface produced from a nadir image and an oblique image collected by aerial cameras popularly utilized in Japan. The results of the analysis indicate that a texture image mapped on a vertical surface produced from a nadir image collected in the usual procedure of aerial photogrammetry would be unable to satisfy the required resolution for the detection of doors and windows. On the contrary, the results indicate that a texture image mapped on a vertical surface produced from an oblique image can be utilized for not only investigation of stories but also detection of doors and windows. However, there remain some problems to solve in the utilization of an oblique image for texture mapping.