The fast drainage of surface water from road pavements is essential to ensure both driving safety and adequate infrastructure service life. For close-graded asphalt mixtures, surface runoff relies on sufficient longitudinal and transverse slopes that convey water toward hydraulic drainage devices. However, construction defects, surface distress, or inadequate placement of drainage systems may compromise this process and reduce pavement durability. When water infiltrates beneath the wearing course and saturates the underlying layers, heavy traffic loads can accelerate deterioration through erosion, pumping, interlayer delamination, and subgrade overstress. This work investigates the joint use of Ground Penetrating Radar (GPR) and Mobile Laser Scanning (MLS) to evaluate drainage deficiencies and detect signs of layer delamination in bituminous pavements. A highway section in Salerno (Italy) was selected as a case study due to known hydraulic-related issues. MLS data were used to reconstruct pavement geometry and model surface runoff patterns, while GPR surveys assessed the condition of the bonding between asphalt and base layers. The results revealed ineffective runoff management and identified multiple areas affected by delamination, confirming a relationship between surface drainage behaviour and subsurface damage. These findings highlight the broader potential of the integrated GPR–MLS framework as a scalable and transferable approach for proactive drainage assessment and structural monitoring in pavement management practices.

Integrating existing tree-level information into harvester operator decision-making can significantly enhance precision forest management, particularly with respect to biodiversity preservation and climate-smart adaptation. During harvester operations, a primary challenge lies in positioning the machine with sufficient accuracy in real time to relate a priori individual-tree-level reference information to the operator. We propose a lightweight procedure using tree-to-tree matching to continuously register a real-time tree map collected from a harvester (or another mobile laser scanning system) to a precomputed reference map derived from an airborne laser scanner (ALS). We assess the robustness of the method using simulated tree maps and validate its real-world performance in experiments using a LiDAR-equipped harvester performing a thinning operation in a boreal forest. In simulations, registration was found to be robust up to a moderate tree density of approximately 1700 ha−1, even when using a reference map with a lower positional accuracy and higher error rates than in our harvester experiments. Using real-world data from the thinning operation, the registration method was demonstrated to successfully mitigate meter-scale positioning drifts remaining in the LiDAR-inertial trajectory. After the continuous registration procedure, the positioning error was reduced to the level of 0.5 m, constrained by the accuracy of the prior map derived from sparse ALS data with ∼5 transmissions/m2. Importantly, the registration procedure was shown to update in real time (at most 20 ms update time for stands with densities of at most 2000 ha−1, after an initial computational phase. Notable features of the registration procedure are its low memory consumption, fast runtime and capacity to accurately position the harvester despite LiDAR-inertial positioning drift. While these results demonstrate the potential for real-time operation, full implementation requires the development of real-time tree detection and estimation of tree attributes.

Urban trees represent valuable renewable biomass sources, but traditional allometric equations inadequately capture structural variability in urban environments. Therefore, considering tree structure is crucial for accurate biomass measurement. While LiDAR-based 3D modeling can reflect tree architecture, generating reliable models from dense foliage point clouds has remained difficult. To address this challenge, we developed the Skeleton Generative Method (SkeletonGM), which reconstructs tree trunks and primary branches from mobile laser scanning data even under heavy foliation. SkeletonGM produces a clarified skeletal point cloud that is subsequently converted into 3D tree models using AdTree; aboveground woody biomass is then calculated by combining the estimated volume with species-specific wood density. To validate the proposed method, we applied SkeletonGM to 33 hinoki cypress trees in a forest stand and 25 dawn redwood trees in a park, and compared the resulting biomass estimates with values derived from allometric equations. The results showed strong agreement with the reference equations (R² = 0.95 and 0.94; mean absolute percentage error = 14.1% and 14.6%). These findings indicate that the proposed method has strong potential to improve the accuracy of biomass assessment for urban trees.

Abstract Mobile laser scanning (MLS) provides detailed point cloud reconstructions of forest environments and has potential for operational forest sample-plot surveying. This study evaluated the accuracy of MLS in deriving forest inventory attributes, including basal area (G), number of trees per hectare (TPH), total stem volume (V), basal area-weighted mean tree diameter (Dg) and height (Hg), and dominant height (Hdom). Experiments were conducted in managed boreal forests across 44 sample plots (370–2000 m2) using a Faro Orbis MLS system. Field measurements collected tree-by-tree (n = 4472) with callipers and clinometers during the previous summer served as reference data. We compared two alternative MLS data acquisition trajectories—closed loops (MLS-loop) and line transects (MLS-line)—and two processing workflows: (i) manually assisted tree detection followed by automatic tree measurements, and (ii) a fully automatic workflow. MLS-line provided similar or marginally improved accuracy compared with MLS-loop; however, the substantially shorter acquisition time of MLS-loop (19.0 min per plot on average) favoured its operational use over MLS-line (30.5 min). Clearer differences emerged between processing workflows. The fully automatic workflow identified and measured 74.1% of trees with diameter at breast height (DBH) > 5 cm, whereas manual assistance in tree detection increased this proportion to 97.1%. DBH accuracy was similar for both workflows (root-mean-square-error [RMSE] ≈ 2.4 cm), but tree-height estimates were substantially less accurate under automatic processing (RMSE 6.2 m) than under the assisted workflow (RMSE 2.1 m). These differences propagated to plot-level estimates. Using the automatic workflow, RMSEs were 4.2 m2/ha for G, 610 trees/ha for TPH, 29.3 m3/ha for V, 2.3 cm for Dg, 1.6 m for Hg, and 1.9 m for Hdom. The assisted workflow notably improved accuracy, yielding RMSEs of 3.5 m2/ha for G, 54.0 trees/ha for TPH, 20.2 m3/ha for V, 1.2 cm for Dg, 1.3 m for Hg, and 1.2 m for Hdom when using closed-loop trajectories. Overall, the results emphasize the importance of assisted workflows for attributes sensitive to detection completeness, particularly TPH, while showing that kinematic MLS can efficiently capture forest structure for sample plot measurements.

• Small greenspaces can buffer extreme heat stress by up to 16 °C mPET on hot days • Vegetation structural complexity is a key driver of thermal comfort in small parks • Vegetation structure affects mPET directly and indirectly via sky view factor • Mature trees with tall canopies provide the strongest mPET buffering in small parks • Shading drives fine-scale thermal variability Urban greenspaces play an increasingly important role in urban planning and public health, particularly in providing thermal comfort during hot summer days. While the cooling potential of greenspaces generally increases with size, it also strongly depends on vegetation structure, particularly in small greenspaces. Optimizing these spaces for thermal comfort requires a clear understanding of how their vegetation structure shapes local cooling. Based on field measurements in 2024, we modeled the modified Physiologically Equivalent Temperature (mPET) across 12 structurally diverse small greenspaces (< 2 ha) in central Munich, Germany, and in their built surroundings. High-resolution vegetation structure derived from mobile terrestrial laser scanning was combined with sky view factor estimates from hemispherical photographs and micro-meteorological measurements. Using mixed-effects models and structural equation modeling, we assessed both direct and indirect pathways linking vegetation structure, sky openness, shading, and thermal comfort. Our results show that small greenspaces can reduce mPET by up to 16 °C compared to built surroundings. Vegetation structure emerged as a key determinant of this buffering capacity: greenspace plots with tall, multi-layered vegetation provided strong mPET buffering, whereas sparsely vegetated plots offered little to no thermal relief. Mean canopy height was the strongest predictor of mPET buffering, and vegetation structure influenced thermal buffering both directly and indirectly through reductions in sky view factor. At finer spatial scales, shading dominated local thermal variability, with shading effects being strongest in structurally complex plots. These findings highlight that small but structurally complex greenspaces can play a vital role in climate-adaptive urban planning.

This paper proposes a novel dataset that has been specifically designed for 3D semantic segmentation of bridges and the domain gap analysis caused by varying sensors. This addresses a critical need in the field of infrastructure inspection and maintenance, which is essential for modern society. The dataset comprises high-resolution 3D scans of a diverse range of bridge structures from various countries, with detailed semantic labels provided for each. The initial objective is to facilitate accurate and automated segmentation of bridge components, thereby advancing the structural health monitoring practice. To evaluate the effectiveness of existing 3D deep learning models on this novel dataset, a comprehensive analysis of three distinct state-of-the-art architectures is conducted. Additionally, data was acquired through various sensors to quantify the domain gap resulting from sensor variations. The findings indicate that all architectures demonstrate robust performance on the specified task. However, the domain gap can potentially lead to a decline in the performance of up to 11.4% mIoU. Code and data are available at https://github.com/mvg-inatech/3d_bridge_segmentation . • The introduction of an enriched dataset of bridges collected with Terrestrial Laser Scanners and Mobile Laser Scanners to serve as a uniform benchmark. • An investigation into the domain gap in 3D semantic segmentation, which arises from the use of disparate data capture sensors. • A general comparative evaluation of the performance among three state-of- the-art architectures in the bridges semantic segmentation task.

Mobile Laser Scanning (MLS) integrated with Simultaneous Localization and Mapping (SLAM) has emerged as a key technology for digitizing GNSS-denied environments, such as underground mines. However, the automated interpretation of unstructured, high-density point clouds into semantic engineering models remains challenging due to extreme geometric anisotropy in point distributions and severe class imbalance inherent to narrow tunnel environments. To address these issues, this study proposes a highly automated scan-to-model framework for precise semantic segmentation and vectorized two-dimensional (2D) profile reconstruction. First, an enhanced hierarchical deep learning network tailored for point clouds is introduced. The architecture incorporates a context-aware sampling strategy with an expanded receptive field of up to 10 m to preserve axial continuity, coupled with a spatial–geometric dual-attention mechanism to refine boundary delineation. In addition, a composite Focal–Dice loss function is employed to alleviate the dominance of wall points during network training. Experimental validation on a field-collected dataset comprising 16 mine tunnels demonstrates that the proposed model achieves a mean Intersection over Union (mIoU) of 85.15% (±0.29%) and an Overall Accuracy (OA) of 95.13% (±0.13%). Building on this semantic foundation, a robust geometric modeling pipeline is established using curvature-guided filtering and density-adaptive B-spline fitting. The reconstructed profiles accurately recover the geometric mean surface of the tunnel wall, yielding an overall filtered Root Mean Square Error (RMSE) of 4.96 ± 0.48 cm. The proposed framework provides an efficient end-to-end solution for deformation analysis and digital twinning of underground mining infrastructure.

Surface defects such as depressions, heaving, and irregular undulations frequently develop on aging concrete bridge decks under repeated traffic loading and environmental effects. Accurate and objective identification of such defects is essential for structural serviceability and safety, yet manual inspection remains labor-intensive and subjective. This study develops a systematic framework for surface defect identification through geometric feature augmentation with a streamlined point cloud learning strategy. In practical engineering scenarios, point cloud data of concrete bridge decks can be periodically acquired via vehicle-mounted mobile laser scanning (MLS) systems and subsequently streamlined for analysis. The proposed method heightens defect sensitivity by extracting interpretable geometric descriptors, further integrating multi-scale representations to capture surface defects across varying spatial extents. Evaluated on a public point-level annotated benchmark, the proposed method clearly outperforms the same network trained with geometric coordinates only. To improve result reliability, all experiments were repeated four times with different random seeds, and the performance is reported as mean ± standard deviation. Results show that the proposed method achieves a precision of 0.597 ± 0.021 and an accuracy of 0.933 ± 0.009 under the benchmark protocol. Overall, these results demonstrate a reproducible proof of concept under controlled benchmark conditions for bridge deck surface defect segmentation, while broader cross-site and cross-sensor validation will be pursued in future work.

ABSTRACT Accurate and scalable estimation of tree trunk diameter, particularly diameter at breast height (DBH), plays a key role in forest inventory, biomass modelling and ecological monitoring. While Terrestrial Laser Scanning (TLS) remains a gold standard for tree structure measurement, the increasing accessibility of wearable and mobile LiDAR platforms raises the need for unified, platform-independent processing strategies. However, most existing workflows are platform-specific and do not provide a standardised basis for comparative evaluation. This study introduces an automated and transferable pipeline for tree trunk segmentation and DBH estimation, designed to work consistently across three distinct LiDAR systems: TLS, Wearable Mobile Laser Scanning (WMLS) and iPad-based LiDAR Scanning (iPLS). Unlike prior studies that rely on manual segmentation or platform-tied algorithms, the proposed method integrates height thresholding (0.5 m), spatial clustering (DBSCAN), trunk validation (PCA) and least-squares circle fitting at 1.30 m to isolate and quantify tree stems. The workflow was tested on 18 field-surveyed trees in a semi-structured forest plot, with reference DBH values used for evaluation. TLS delivered the highest accuracy (MAE = 0.92 cm, RMSE = 1.10 cm, R2 = 0.98), followed by WMLS (MAE = 2.15 cm, RMSE = 2.71 cm, R2 = 0.92), which balanced portability with precision. iPLS showed reduced performance (MAE = 3.98 cm, RMSE = 4.42 cm, R2 = 0.80) due to limited sensing range and data density. Additional issues included SLAM-induced trunk deformation in WMLS and segmentation artefacts in iPLS. These findings demonstrate that geometry-based, modular DBH estimation can be effectively applied across LiDAR platforms with varying resolution and mobility. The proposed framework fills a gap in cross-platform forestry workflows by enabling objective, reproducible and sensor-agnostic trunk measurements suitable for both research and applied inventory applications.

The preservation of built heritage is an essential practice, as these assets embody historical, cultural, and technical values whose degradation or loss irreversibly compromises collective memory. Buildings of historical interest require continuous inspections and monitoring that, in many cases, must not cause alterations or damage to their structure. In this context, this article aims to investigate techniques for monitoring the structural health of buildings of historical interest. To this end, the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) method was employed, guiding the selection and analysis of scientific articles and enabling the development of a structured systematic literature review (SLR). The research focused on non-destructive testing methods and on digitization techniques and digital twin creation, which are directly related to studies on the characterization and documentation of built heritage. The results revealed that visual inspection, Ground Penetrating Radar (GPR), and Infrared Thermography (IRT) are widely used for the characterization of historic buildings. Digitization for the creation of digital twins in built heritage has been carried out through photography, photogrammetry, and mobile laser scanning, with the aim of monitoring deterioration processes, among other contributions to the field of conservation.

A three-stage framework for stand-level automated stem volume estimation in temperate forests using Mobile laser scanning

Abstract. Historic gardens are living forms of Cultural Heritage whose spatial identity is inseparable from continuous processes of growth, decay, and maintenance. Although recent advances in laser scanning, photogrammetry, and mobile mapping systems enable highly accurate three-dimensional documentation, most digital models remain limited to static representations. The temporal dimension, essential for understanding and managing garden heritage, is rarely integrated as an intrinsic component of spatial data. This paper explores an experimental four-dimensional integration framework for historic gardens that combines point cloud data, semantic and multi-source data fusion based on GIS, and web-based 4D visualization. Rather than aiming at a complete temporal reconstruction, the approach investigates how a single-epoch 3D survey can act as a temporal anchor for integrating historical documentation and future-oriented scenarios within a unified spatial environment. The framework is tested on the historic garden of Villa Burba (Rho, Milan, Italy). Using open-source tools, point clouds from mobile laser scanning are processed using machine learning and semantically structured in a GIS environment, where time is modelled as a relational property describing transformation processes. The integrated model is visualized in the Cesium web platform, enabling interactive exploration of spatial-temporal relationships. The results demonstrate the feasibility of a scalable and interpretable 4D framework for living landscape heritage.

Point cloud density approach to characterize and estimate shrub fuel load in the mediterranean environments using mobile laser scanning

Abstract 3D thermal models are associated with building inspection and energy efficiency evaluation. Fusing Thermal infrared (TIR) images with MLS (Mobile Laser scanning) point clouds enables the generation of thermal point clouds, which combine detailed geometric data with thermal attributes at each 3D point. RGB images are typically used to reconstruct a 3D point cloud and apply thermal textures to the model. Therefore, the generated thermal point cloud heavily relies on accurate RGB reconstruction and scale estimation. In this contribution, we introduce a novel image-feature alignment method to directly co-register TIR images with MLS point clouds. The intensity images are generated from the point clouds, and corresponding feature points are matched with the TIR images. With the estimated corresponding points, the pose can be calculated, and the thermal textures are projected onto the MLS point clouds for thermal point cloud generation. Our method achieves results comparable to manual labeling with a projection error of RMSE 3.4 pixels, offering an efficient and reliable solution to generate 3D thermal models for building energy evaluations.

Due to the underground or enclosed nature of subway systems, satellite navigation signals are often difficult to receive. Without reliable signals, mobile measurement systems are prone to significant positioning errors. To address this challenge, this study proposes an improved method that leverages control points, dynamic calibration, and advanced data fusion. A unified coordinate system is established for all sensors, ensuring that measurements from different devices, such as inertial navigation systems and laser scanners, can be integrated seamlessly. System errors are identified and corrected through dynamic calibration to maintain long-term accuracy. Object detection technologies are used to identify subway control points, enabling trajectory correction through coordinate-assisted updates. The calibrated parameters are then applied to fuse inertial navigation data with laser scanning data, effectively reducing the accumulation of errors. This approach generates high-precision three-dimensional point clouds with absolute coordinates, suitable for applications such as subway inspection and tunnel monitoring in signal-denied environments. Results show that the method achieves deviations within 1 cm for horizontal measurements and 5 mm for vertical measurements. This significant improvement enhances the precision and reliability of laser scanning data in underground applications. The study also analyzes the spatial distribution of control points on point cloud quality, providing a practical solution for improving mobile measurement systems where satellite signals are unavailable.

The box-counting method is sensitive to tree positions in the coordinate system, and based on this method, assuming self-similarity of beech tree models derived from mobile laser scanning was difficult. The repetitive branching architectures of trees lead us to think that trees exhibit self-similarity and fractal properties. Therefore, applying fractal analysis to assess tree structures may seem natural. This study aimed to evaluate the box-counting method (BCM), a simple and commonly used fractal analysis, for describing 3D biological trees, using a wide range of structural models of European beech trees (Fagus sylvatica L.) derived from mobile laser scanning. We specifically investigated the method’s sensitivity to the arbitrary placement of a tree within the coordinate system and the validity of the tree’s self-similarity based on the BCM. The BCM was sensitive to the tree position in the coordinate system, with observed minimum and maximum variations in the box-counting dimension (Db) values of 0.18 and 0.52, respectively, when translating the trees in the XYZ directions. The analysis of the local slopes of the BCM, which are the slopes between neighboring data point pairs, revealed a large variation of slope values across scales with a clear pattern, indicating that the structural patterns of the sampled trees were inconsistent and locally dependent. Thus, it is difficult to assume self-similarity of the beech tree models based on the BCM. Our results demonstrate the need to standardize the computation of the Db for single trees with respect to the coordinate system and caution in interpreting the Db. These findings contribute to a deeper understanding of the Db to assess tree structures and functions.

The installation quality and long-term stability of roadway roof bolts/cables are critical to the safety of deep coal mines, yet conventional manual inspections are time-consuming, subjective, and difficult to perform frequently. This paper presents a mobile laser scanning (MLS)-based method for automatic recognition and parameter extraction of roof bolts/cables from 3D point clouds. A SLAM-based MLS system is first used to obtain a continuous 3D model of the roadway. The roof is then globally aligned and separated from surrounding objects using a cloth simulation filter (CSF), which yields candidate roof points. Instance-level clusters are extracted by density-based clustering combined with geometric constraints, and principal component analysis (PCA) is applied to derive key geometric parameters for each bolt/cable, including exposed length, inclination, and row/column spacing. Field experiments were carried out on five consecutive segments of a return airway in a deep coal mine, with 127 manually labeled bolts/cables. The proposed method correctly identified 118 of them, achieving a precision of 96.72% and a recall of 92.91%, while also providing an automatically generated parameter database for engineering evaluation. The results indicate that the method can effectively support intelligent assessment of roadway support quality, although its performance remains dependent on MLS data quality and the visibility of exposed bolts/cables, highlighting the need for further validation in additional mines and under more extreme monitoring conditions.

Abstract. Tree volume estimation is fundamental to forest management and inventory, yet traditional methods rely on allometric equations that introduce significant uncertainties due to generalized relationships and measurement limitations. This study evaluates the accuracy of mobile personal laser scanning (PLS) technology for tree volume estimation in pedunculate oak (Quercus robur L.) forests through a controlled comparison framework. Field work was conducted in January 2025 under optimal leaf-off conditions in a lowland oak stand in central Croatia. Three morphologically typical mature oak trees were selected within a single plot to enable controlled comparison while minimizing environmental variability. Data was acquired using PLS Faro Orbis Scanner, emphasizing complete stem coverage from multiple azimuths to support robust SLAM trajectory estimation and minimize occlusion effects. Three principal volume estimation approaches were evaluated: (i) sectioning volume obtained after felling, serving as the operational reference; (ii) PLS Schumacher-Hall volume, computed from LiDAR derived DBH and total height using established allometric relationships; and (iii) PLS Trunk Volume, computed directly from point cloud data using LiDAR360's trunk slicing workflow. Following PLS data acquisition, target trees were felled and bucked into contiguous sections, with length and end diameters recorded for each section to compute reference volumes. The sectioning dataset was treated as an operational reference rather than absolute ground truth, acknowledging potential reconstruction errors due to field conditions and occasional stem breakage. The study reveals important trade offs between measurement accuracy, operational efficiency, and methodological complexity, with sectioning volume providing the most direct measurement approach by eliminating remote sensing processing uncertainties. The research establishes a robust methodological framework for evaluating PLS performance in oak forests while highlighting both significant potential and current limitations of mobile laser scanning for operational forest inventory applications.

Abstract. The quality assessment and updating of spatial geodatabases (geoDBs) are essential tasks for effective spatial data management. This paper introduces an innovative methodology called Virtual Reconnaissance (VRec), which leverages Mobile Laser Scanning (MLS) systems based on Simultaneous Localization and Mapping (SLAM) technology. VRec aims to support both field reconnaissance and the geometric/semantic validation of GeoDBs. After reviewing the state-of-the-art, a case study in the municipality of Lecco (Italy) is presented, where a portable MLS device was used to acquire high-resolution point clouds. These data were georeferenced using GNSS ground control points (GCPs) and compared with the existing geoDB. Results demonstrate that VRec enables accurate quality assessment within official tolerance thresholds and offers promising capabilities for GeoDB updating, especially in complex urban environments. While data processing still requires skilled operators and significant time investment, future integration with artificial intelligence techniques may enhance efficiency and scalability.

A 3D point cloud is a collection of millions of spatial points that faithfully capture object shape and structure. Handheld mobile laser scanners are increasingly adopted because they enable rapid and complete acquisition. This paper evaluates the geometric and metrological quality of a point cloud collected with a handheld MandEye PRO scanner based on the Livox Mid-360 from an architectural object, using terrestrial laser scanning with a Leica ScanStation P40 and a network of total station control points as references. Point-like, linear, and planar features were assessed by registration to the control frame, segment-length comparisons, plane fitting, and cloud-to-cloud distance analysis. After registration, root mean square errors on control points fall within 0.02–0.19 m (mean 0.08 m). Segment lengths are consistent to within a single centimetre across handheld MLS, TLS, total station measurements, and direct field readings, while the differences in areas of fitted planes are 0.7–1.9%. Local deviations concentrate along edges and in shadowed zones, indicating sensitivity to trajectory coverage and sampling density. In well-covered regions, the overall agreement between MLS and TLS remains stable, whereas gaps in visibility or sparse sampling lead to localised discrepancies. The results show that, in this configuration, handheld mobile scanning provides accuracy consistent with the requirements of geodetic documentation while offering high acquisition efficiency. These findings support the use of handheld MLS for architectural surveying and geodetic fieldwork, provided that route planning and sampling are designed to ensure robust coverage of critical facades, edges, and occluded areas.