Abstract. The demand for accurate, lightweight 3D building models is rapidly growing in urban analysis, digital twins, and geospatial information systems. Single-source airborne point clouds, such as airborne laser scanning (ALS) or dense image matching (DIM), often suffer from geometric incompleteness, uneven density, and misalignments, limiting the reliability of Level of Detail (LOD) building reconstructions. While substantial progress has been made in single-source building reconstruction and multi-source fusion, fully automated LOD generation pipelines that effectively exploit cross-source airborne data remain limited. This paper presents an automated workflow for generating precise LOD building models from cross-source fused point clouds, leveraging the precision of ALS and the high resolution of DIM to improve model fidelity. Using point clouds obtained from a slice-to-slice fusion approach, experiments on Luxembourg datasets demonstrate a reduced model standard deviation of 0.17m compared to 0.20m for ALS, 0.29m for DIM, and 0.27m for conventional ICP-based fused point clouds. The results show that our workflow, combined with a polygon fitting algorithm and cross-source fused data, significantly enhances building model accuracy and geometric completeness, highlighting the value of multi-source integration for automated 3D city modeling.

BIM Adoption in firms and projects requires considerable changes in design and construction processes. There has been ongoing research on exploring the drivers and barriers to BIM adoption. BIM Tools can be defined as all software tools and applications that can input information into/acquire information from semantically rich digital 3D building models, also known as Building Information Models (BIM). The aim of this study was to identify the impact of demographic, social, education-related, previous training-related, and profession-related factors on the perception of the ease of use and usefulness of BIM Tools for students and early career professionals. The main question of the research was defined as follows: “What factors influence the perception of the ease of use and usefulness of BIM tools for students and early career professionals?” The study involved a questionnaire survey with 227 participants to measure the impact of eleven different factors on the perception of the ease of use and usefulness of these tools. The findings suggest that both the perceived ease of use and perceived usefulness of BIM Tools are mostly stable and not substantially affected by most of the external factors. Among the factors that can influence the perceived ease of use and perceived usefulness, Participation in a BIM Certification Training Program appeared to be the factor with the strongest influence, as it had a significant influence on both dimensions. Factors with weaker influence included Age Group, Gender, Being a Student or Not, Computing Habits, and Gaming Habits. The other five factors investigated appeared to have no influence on either dimension.

The influence of rolling stock loading on a twenty-three-story frame building located near the movement of railway trains in an urban area was investigated. Mathematical modeling of the dynamic behavior of multi-story buildings subjected to rolling stock loading was performed using a two-stage numerical method. In the first stage, a finite element model of the multilayer soil foundation along with the ballast prism was created in the NASTRAN software complex, represented as a planar elastoplastic half-space with a length of 200 m and a depth of 30 m. A real geological cross-section consisting of five layers with different physical characteristics was used. The rolling stock load is represented as a vertical periodic excitation, concentrated at the center of mass of the system consisting of the bogie frame, the wheelsets of a freight wagon, and the ballast prism. Modal analysis of the soil foundation and the ballast prism was performed using the Lanczos method. The influence of rolling stock loading on the dynamic behavior of the soil foundation was investigated using the fourth-order Runge-Kutta method. Horizontal and vertical displacements and accelerations of the soil were obtained at various distances and depths of the foundation model from the railway track axis. In the second stage, a 3D model of the monolithic frame building was created in the SCAD software complex. Modal analysis of the structure was performed using the subspace iteration method. Two calculation options for the multi-story building were considered. The first calculation was performed for the action of design load combinations: permanent, sustained, and short-term (snow, wind load). In the second calculation option, the stress-strain state of the building was investigated using the spectral method under the action of design loads and kinematic soil excitation, applied along the height of the building's foundation in the form of acceleration vectors. The accelerations were considered in two directions and added to the design combinations along the two directions of wind load influence. A comparison of the two calculation options was performed to check the reliability and structural safety of the building.

Over the past couple of decades, technology has become an integral part of our lives, with widespread digitalization and access to knowledge making our lives easier every day. What seemed like science fiction in the 20th century is now perceived as completely ordinary and mundane. One such example is unmanned aerial vehicles (UAVs). Today, UAVs are used in many areas: agriculture, construction, energy, and military applications. UAVs are a very promising field in geodesy and cartography. In vast and hard-to-reach areas, UAVs reduce time, money, and labor costs. Using specialized software, it is possible to create both highly accurate terrain maps and 3D models of buildings and structures. This article presents the experience of creating a 3D model of a residential building in the city of Yakutsk. Currently, with the possible introduction of a 3D real estate cadastre and information modeling technologies in construction, this research topic is highly relevant.

Large-scale 3D point cloud modeling plays a key role in building recognition and the construction of digital cities. However, existing methods generally suffer from low computational efficiency, insufficient clustering accuracy, and unstable modeling results. Therefore, this study proposes a 3D building modeling approach using Density-Based Spatial Clustering of Application with Noise. The method introduces geodesic distance and neighborhood search strategies, combined with a hierarchical coordination grid partitioning mechanism that considers boundary curvature features, to improve modeling accuracy and processing efficiency. Experimental results show that the model achieves an area under the curve, harmonic mean of precision and recall, standard mutual information, Jaccard index, and adjusted Rand index of 0.984, 0.957, 0.973, 0.9754, and 0.978, respectively, all significantly higher than those of the comparison models. At the same time, the model’s point cloud restoration accuracy, as measured by the root mean square error and mean absolute percentage error, is 4.9% and 6.1%, both of which are significantly lower than those of the three comparison models. These results indicate that the proposed 3D modeling model demonstrates strong advantages in clustering performance, spatial structure restoration ability, and operational efficiency. It shows great scalability and practical value, providing solid technical support for urban digital construction and spatial information reconstruction.

Integrated physics-machine learning for real-time urban photovoltaic mapping: Coupling local climate zones with 3D building models

Digital surface models (DSMs) derived from high-resolution satellite imagery often contain mismatches, voids, and coarse building geometry, limiting their suitability for accurate and standardized 3D reconstruction. The scarcity of finely annotated samples further constrains generalization to complex structures. To address these challenges, an automated building reconstruction method based on two-stage polygon decomposition and adaptive roof fitting is proposed. Building polygons are first extracted and standardized to preserve primary contours while improving geometric regularity. A two-stage decomposition is then applied. In the first stage, polygons are coarsely decomposed, and redundant rectangles are removed by analyzing containment relationships. In the second stage, non-flat regions are identified and further decomposed to accommodate complex building connections. For 3D model fitting, flat-roof buildings are reconstructed by integrating structural analysis of DSM elevation distributions with adaptive rooftop partitioning, which enables accurate modeling of complex flat structures with auxiliary components. For non-flat roofs, a representative parameter space is defined and explored through systematic search and optimization to obtain precise fits. Finally, intersecting primitives are normalized and optimally merged to ensure structural coherence and standardized representation. Experiments on the US3D, MVS3D, and Beijing-3 datasets demonstrate that the proposed method achieves higher geometric accuracy and more standardized models, with an average IOU3 of 91.26%, RMSE of 0.78 m, and MHE of 0.22 m.

Three-dimensional (3D) building models are essential for urban planning, spatial analysis, and virtual simulations. However, most reconstruction methods based on Airborne LiDAR Scanning (ALS) rely primarily on rooftop information, often resulting in distorted footprints and the omission of façade semantics such as windows and doors. To address these limitations, this study proposes an automatic 3D building reconstruction method driven by façade geometry. The proposed method introduces three key contributions: (1) a façade-guided footprint generation strategy that eliminates geometric distortions associated with roof projection methods; (2) robust detection and reconstruction of façade openings, enabling reliable identification of windows and doors even under sparse ALS conditions; and (3) an integrated volumetric modeling pipeline that produces watertight models with embedded façade details, ensuring both structural accuracy and semantic completeness. Experimental results show that the proposed method achieves geometric deviations at the decimeter level and feature recognition accuracy exceeding 97%. On average, the reconstruction time of a single building is 91 s, demonstrating reliable reconstruction accuracy and satisfactory computational performance. These findings highlight the potential of the method as a robust and scalable solution for large-scale ALS-based urban modeling, offering substantial improvements in both structural precision and semantic richness compared with conventional roof-based approaches.

The quality of teaching and learning activities is a key indicator of success in formal education, and the provision of comfortable classrooms plays a vital role in supporting optimal learning outcomes. This community service aimed to analyze and improve the thermal comfort of the proposed classroom design at MTs Ummu Aiman Malang, ensuring that the building meets thermal comfort standards and regulations. A 3D model of the school building was developed, and a Computational Fluid Dynamics (CFD) simulation was employed to assess indoor air distribution and airflow. Based on the theoretical analysis and simulations conducted, some alternative design optimizations suggested are choosing the color of the exterior wall paint with a light color gradation that has a good heat absorption value, using a type of double-glaze glass that has better thermal performance, adding heat insulation material to the exterior walls of the east and west sides, and modifying the opening of the window to maximize natural ventilation and improve air circulation in the classroom. By optimizing the design, the OTTV value of the MTS Ummu Aiman building was reduced from 97.07 W/m² to 33.49 W/m². In addition, the addition of openings in the window changes the temperature range in the classroom, which was previously 26.6 -29.3 °C, to a temperature range of 22.8-25.8 °C, with the air velocity falling above the occupants' heads not exceeding 0.25 m/s. The results demonstrate that design optimization and appropriate material selection significantly improve thermal comfort and energy efficiency, providing a practical reference for sustainable school building design.

Digital Twins (DTs) are transforming construction and energy management sectors by integrating 3D surveying, monitoring, Building Performance Simulation (BPS), and Building Energy Simulation (BES) from the earliest design or retrofit stages. Moreover, dynamic thermal simulations further support energy performance assessments by modeling indoor conditions to meet comfort and efficiency targets. However, their reliability depends on accurate, standards-compliant 3D building models, which are costly to create. This research introduces a complete framework for automatically generating energy-focused Digital Twins (EDTs) directly from unstructured point clouds. Combining Deep Learning-based instance detection, Scan-to-BIM techniques, and computational geometry, the method produces simulation-ready models without manual intervention. The resulting EDTs streamline early-stage performance evaluation, enable scenario testing, and enhance decision making for energy-efficient retrofits, advancing smart-building design through predictive simulation.

Abstract Three-dimensional modelling of buildings requires reliable data sources and sophisticated tools capable of delivering exhaustive models that can facilitate property management. The authors devised a methodology for 3D building modelling using only open-access geospatial databases like Light detection and ranging (LiDAR) scanning, map and photogrammetry resources, and the relevant publicly available land and topography databases. The models integrate building geometry and detailed object information, which makes them versatile tools for property valuation, management, and structural health monitoring. The study brings together Building Information Modeling (BIM) and Geographic Information Systems (GIS) tools that can integrate spatial data and build precise models with details of critical parts of buildings, such as roofs, walls, window and door openings, balconies, terraces, hard infrastructure, and other structural and fit-out components. The methodology’s performance and versatility were verified on single-family residential buildings in Kraków (Poland). The results have confirmed that the constructive collaboration of open-access geospatial data, GIS, and BIM yields high-grade 3D models for structural health monitoring, action planning, and building life cycle management. This approach leads to effective property (resources) management and streamlines planning and taking actions over the life cycle.

Abstract. Feature matching is a necessary step for many computer vision and photogrammetry applications such as image registration, structure-from-motion, and visual localization. Classical handcrafted methods such as SIFT feature detection and description combined with nearest neighbour matching and RANSAC outlier removal have been state-of-the-art for mobile mapping cameras. With recent advances in deep learning, learnable methods have been introduced and proven to have better robustness and performance under complex conditions. Despite their growing adoption, a comprehensive comparison between classical and learnable feature matching methods for the specific task of semantic 3D building camera-to-model matching is still missing. This submission systematically evaluates the effectiveness of different feature-matching techniques in visual localization using textured CityGML LoD2 models. We use standard benchmark datasets (HPatches, MegaDepth-1500) and custom datasets consisting of facade textures and corresponding camera images (terrestrial and drone). For the latter, we evaluate the achievable accuracy of the absolute pose estimated using a Perspective-n-Point (PnP) algorithm, with geometric ground truth derived from geo-referenced trajectory data. The results indicate that the learnable feature matching methods vastly outperform traditional approaches regarding accuracy and robustness on our challenging custom datasets with zero to 12 RANSAC-inliers and zero to 0.16 area under the curve. We believe that this work will foster the development of model-based visual localization methods. Link to the code: https://github.com/simBauer/To_Glue_or_not_to_Glue

The development of geospatial technology has created a need for three-dimensional (3D) spatial data to meet various requirements. 3D models are considered capable of accurately representing the real conditions of an object, making them a vital tool in the planning process. This research aims to analyse the results of integrating point cloud data from Terrestrial Laser Scanners (TLS) and Airborne Laser Scanners (ALS) in creating a 3D model of the Rectorate Building - ITS. The results of data acquisition and processing from the Terrestrial Laser Scanner yielded a mean bundle error value of 0.004 meters, demonstrating a high level of accuracy. Additionally, the results from GCPs showed RMSE_x of 0.005 meters, RMSE_y of 0.013 meters, and RMSE_z of 0.009 meters. According to the ALS data, the average altitude difference is 0.037 meters, and the RMSE is 0.048 meters. The integration of TLS and ALS point cloud data enables the representation of the object’s geometry as a whole, encompassing both its exterior shape and interior structure. The registration results show an RMSE_x value of 0.012 meters, an RMSE_y value of 0.015 meters, and an RMSE_z value of 0.008 meters, further confirming the high level of accuracy. This corresponds to a high level of accuracy and is based on the three-dimensional modelling standard of the Open Geospatial Consortium at LOD 4, with a precision of 0.200 meters. The geometric accuracy test results of the 3D model length comparison with the existing building length data show an RMSE value of 0,033 meters, further demonstrating the high geometric accuracy and meeting the Level of Detail (LOD) 4 accuracy standard.

Abstract Generating Building Energy Models (BEM) of Positive Energy Neighbourhoods (PEN) is a challenging task due to the cost of the generation of detailed building models or the inaccuracy or lack of representation of archetype-based models. To solve this problem, in past years, several tools appeared to automatically convert BIM (Building Information Model) data to BEM in order to reduce the burden of creation of the 3D building model. In this work, the available BIM to BEM conversion tools are reviewed highlighting their differences and unique characteristics. This review shows a research gap where there has been advance in the recent years, but no available tools have come out public for this specific purpose. Following available literature we develop a graph-based IFC extraction process using open-source python libraries and test it with a real-world building. To complete the BEM generation process, the extracted BIM information was enriched with non-geometrical data programmatically, obtaining a fully functional model. The resulting model is compared with a standard detailed model created by an experienced engineer to test the reliability of the process. The developed BIM to BEM conversion process has proven to be orders of magnitude faster than available libraries, automatically creating the approximated building geometry and space adjacencies necessary to build a wide variety of energy models with different purposes such as district planning or Model Predictive Control.

Evaluation of 3D building model quality is central to civil engineering, geodesy, architecture, and the construction industry as it facilitates the assessment of the model's accuracy, completeness, and interoperability. The article presents a concept of a versatile method for evaluating the quality of 3D building models, the 3D Model Evaluation Method. It can compare models regardless of their data sources, modelling techniques, visualisation, and file format. The method covers such key quality aspects as completeness, which reflects the degree to which object components are represented; accuracy, which concerns geometric and positioning integrity; and interoperability, which includes the model's compliance with standards, editability, and reusability. The author investigated a range of 3D modelling approaches, including mesh, solid, and parametric models derived from a variety of data sources. The proposed evaluation method is founded on analysing model attribute values for selected criteria followed by visualisation of these values on a radar chart. The model quality index, normalised to [0;1], quantifies the 3D model quality. The results demonstrate the method's effectiveness for 3D model evaluation regarding both geometric and non-geometric aspects. The method can be applied in GIS, BIM, spatial analyses, civil engineering, and environmental engineering. It provides a single pipeline for classifying and comparing different types of models. The 3D Model Evaluation Method provides a universal, structured, and practical basis for comparing 3D models across sources and techniques, ensuring result comparability and consistent quality reporting.

Photogrammetry is a mapping method that utilizes photographic images to generate three-dimensional (3D) models. In this study, the photogrammetry method is used to create a 3D model of the STMIK TIME campus buildings to document and visualize the structures digitally. The process of creating the 3D model begins with image acquisition using a digital camera, followed by processing with photogrammetry software to generate point clouds, meshes, and textured models. The final result of this research is a 3D model of the STMIK TIME campus, which can be used for various purposes such as architectural planning, building preservation, and visual simulation. Based on the evaluation conducted, the resulting 3D model demonstrates a high level of accuracy in representing the actual building structure. Thus, the photogrammetry method has proven to be an effective technique for creating 3D building models at a more affordable cost compared to conventional 3D modeling techniques.

This research discusses the development of the “ARVincent” application, an Android-based Augmented Reality application used to visualize public buildings in 3D using the Marker Based Tracking method. The purpose of this research is to create interactive learning media that can help children recognize public buildings in a more interesting and educational way. The development was carried out using Unity and Vuforia as AR platforms, and SketchUp for 3D object creation. The final result is an application that is able to display a 3D model of a public building when a marker is recognized by the device's camera. Although limited to one object and can only be used offline, this application is proven to provide good visualization and easy interaction for users.

PointNet++ has the functions of object recognition and semantic segmentation, and performs well in the recognition and classification of similar objects. It has been widely used in outdoor 3D scenes. The functional space layout of subway station buildings has the characteristics of similarity and replicability, so it is of great significance to adopt intelligent algorithm to modular design of functional space. In this study, plane data of subway stations in several cities were collected. Build 3D model and export cloud point data (X, Y,Z, rgbC), then enhance the data. The data set is divided into training set, verification set and test set according to the ratio of 8:1:1. The PointNet++ is used to train the 3D data set. Results show that the 3D data set derived from the building model runs well in the PointNet++ network, and can realize the effective transmission of information. Firstly, Pointnet++ model’s recognition and classification results of training set meet expectations, and the stable values of Training mean loss and Training accuracy are 0.42 and 0.76, respectively. Secondly, the training model is used to predict the test set. The Eval avg class Accuracy and Eval Accuracy are 0.80 + and 0.75 + respectively, and the mean intersection over union is about 60%. Thirdly, the prediction results of the test set in the training model show that the Ground Truth and Predict are highly matched, and the model has a strong self-learning ability and can actively optimize the scheme space. This paper proposes a modular design method based on deep learning of 3D spatial data, which can identify and classify architectural space efficiently and accurately.

ABSTRACTPrestack depth migration requires a velocity model that extends from the surface through shallow to deep layers. Generally, the shallow velocity model is built using first‐arrival wave, under which the velocity model is built from reflected wave. These two models are integrated to form a continuous velocity model spanning from the surface to deep layers. Conventional shallow velocity model building methods are restricted by non‐uniqueness of inversion, limited inversion depth and poorly defined reflection interface, which hinders their integration with deeper reflection‐derived velocity models. Errors in the shallow velocity model can compromise the imaging quality of mid‐deep. In this article, we propose a 3D shallow velocity model building method using the joint inversion of first‐arrival and reflection traveltimes constrained by well data. This approach enables simultaneous inversion of velocity and reflection interfaces, thereby facilitating the integration of shallow and mid‐deep velocity models. Combining the first‐arrival and reflected waves enhances ray coverage angles and folds. By constraining the inversion for velocity and reflection interface with logging velocity and drilling depth, respectively, the method effectively mitigates the non‐uniqueness of inversion, thereby improving the accuracy of shallow velocity model. The inverted reflection interface can also be used for integration of shallow velocity model and its underlying velocity model. The theoretical model test proves the feasibility of the method, and the field data application verifies the effectiveness of the method.

Urban morphological analytics on buildings is informative for sustainable development. 3D building massing features, such as courtyards and setbacks, reflect spatial organizations and circulations, while influence daylight access, ventilation, and shading. However, existing 3D GIS methods usually overlook such 3D massing features, further obscure morphological analytics and environmental assessment. This article proposes MorphCut, an efficient convex decomposition method that segments 3D shapes into mass-aligned parts. MorphCut leverages key morphological properties—planarity, regularity, and Gestalt laws—after a topological preprocessing step to enable mass-aware decomposition. Experiments on representative samples, ranging from small houses to complex skyscrapers, showed that MorphCut outperformed four baseline methods in (i) balancing convexity and compactness, (ii) aligning decomposed parts with building masses, and (iii) preserving geometric fidelity (average deviation: 0.25 m). An urban-scale validation on datasets from Delft and Hong Kong, comprising over 30,000 buildings across 18.3 km², demonstrated MorphCut’s robustness, scalability, and generalizability. MorphCut successfully decomposed 98% of buildings in low-rise regions (+78% over the second-best method) and 93% in high-rise areas (+2%), completing processing in 13 hours (3 hours faster). These results position MorphCut as a foundational 3D GIS tool for large-scale, mass-aware morphological analysis, with implications for digital twins, sustainable planning, and environmental modeling.