Are visual surveys outdated? A case study comparing visual surveys designed for sun coral assessment and emerging 3D photogrammetric approaches.

The demand for analyzing images from sources such as closed-circuit television cameras has increased significantly. Conventional analyses, including gait and soft biometrics, typically require the comparison of two video footage clips, as these methods are predicated on video-to-video comparisons. Moreover, numerous prerequisites often limit their applicability, particularly in the field of gait biometrics. To address these limitations, this paper introduces a simple yet effective image-to-person comparison method, leveraging image reproduction from a structure from motion (SfM)/photogrammetry-based three-dimensional (3D) computer graphics reference virtual avatar. This avatar is generated from a reference real person. It is demonstrated that the proposed method, by applying 3D joint manipulations to the reference virtual avatar, qualitatively reproduces a person captured in a target image with high fidelity. Furthermore, quantitative silhouette comparisons successfully confirm distributions for forensic image-to-person comparison. The proposed method holds promise as a body shape-based forensic image-to-person comparison tool in scenarios where a real person can be used as a reference.

3D Gaussian Splatting (3DGS) has been explored for surface reconstruction, however unstructured and discontinuous Gaussian point clouds lead to uneven surface reconstruction accuracy as well as frequent loss of Novel View Synthesis (NVS) quality. To address this problem, we propose a scale-aligned monocular depth-guided 3DGS, a promising novel framework that combines geometric prior regularization and consistency supervision to achieve high-quality rendering and surface reconstruction. Specifically, monocular depth, estimated by some recently proposed monocular depth estimation models, contain implicitly abundant valuable geometric cues, but scale ambiguity limits its application. Therefore we first propose a $K$-Nearest Neighbor (KNN)-based depth alignment framework that utilizes the full-domain gradient at monocular depth map to align to the sparse point cloud obtained during the Structure from Motion (SfM), which is employed for regularization to enhance geometric representation. Then a pseudo-mesh-based multi-view consistency module is introduced to fine-tune and guide the model to recover the accurate surface. Finally, a pixel-level isotropic gradient aware method guides the appropriate growth of the Gaussians to further improve the surface and rendering quality. Experiments on dozens of indoor, outdoor, and object-centered/non-object-centered datasets demonstrate that our method achieves accurate surface reconstruction with excellent NVS performance. The code will be available at https://versewei.github.io/SMDGS/.

The high-fidelity 3D digitization of small, detailed cultural heritage objects, such as Oracle Bones, presents significant challenges for which existing reconstruction workflows are often inadequate. Methods based on Structure-from-Motion (SfM) often lack the geometric density required to capture fine inscription details, while Light Detection and Ranging and RGB-Depth approaches may introduce high data overhead and unstable color mapping. Recent specialized studies have utilized multi-shading-based techniques to extract such hidden surface textures, yet integrating these results into a cohesive mesh remains difficult. To address these limitations, we propose a digitization framework specifically designed for object-level archaeological artifacts. Our method combines semi-automatic alignment with ICP-based refinement for robust camera pose estimation, reducing misalignment issues associated with feature-only registration. Furthermore, we employ an efficient mesh-based representation with vertex-level coloring, enabling detailed geometry and consistent texturing while maintaining compact storage requirements. Our contributions include: (1) a high-quality mesh reconstruction framework that preserves fine inscription geometry; (2) a hybrid camera pose estimation strategy that improves alignment robustness; and (3) an integrated hardware-assisted workflow tailored for digitizing small archaeological artifacts under controlled acquisition conditions. Experimental results on physical Oracle Bone artifacts demonstrate that the proposed method achieves a mean geometric reconstruction error of approximately 0.075 mm with a Hausdorff distance of 1 mm. These results demonstrate the effectiveness of the proposed workflow for digitization of oracle bone artifacts.

Abstract Unoccupied Aerial Vehicles (UAVs) and Structure from Motion (SfM) photogrammetry have revolutionised data capture on the coast. In remote or hazardous areas, they enable the collection of morphological data from areas that are unable to be physically accessed, such as on vertical cliffs where active collapse is occurring. The digital surface models that are produced from vertical photogrammetry can, however, be limited in areas where cliff faces are uncut at their base, or the cliff top overhangs its face. To solve this issue, this study tests the applicability of UAV‐SfM‐derived point clouds for assessing cliff morphological change. A four‐year‐long timeseries, with a bi‐monthly sampling resolution (n = 29), was analysed for a 1.5 km stretch of vertical and overhanging sea cliffs, formed in soft clay of Tertiary age on the open coast of Victoria, Australia. The retreat rate for the upper half of the cliff face was 0.67 m/year (0.60 m 3 /m/yr), with nine high magnitude cliff‐top collapses, exceeding 1,000 m 3 (up to 9,500 m 3 ), occurring in the study period. Pre‐collapse deformation, namely seaward tilting of the face, was detected prior to 75% of collapses > 300 m 3 . Deformation was observed to occur in two ways. The first, preceding most large collapses (> 500 m 3 ), was caused by the expansion of tension cracks behind the cliff top. The second, associated with smaller collapse volumes (100–500 m 3 ), was initiated by rock slabs fracturing and cleaving away from the cliff face. An additional 14 instances of seaward displacement of cliff face have been identified that have not yet resulted in collapse. This study highlights the benefits and potential for using UAV‐SfM‐derived point clouds for the monitoring of hazardous cliff environments. Benefits extend from ease of data capture and generating extensive time series to the analytical insights it can provide. UAV SfM point clouds offer a promising low‐cost alternative to cliff monitoring compared to commonly used techniques such as Terrestrial Laser Scanning (TLS), especially in difficult‐to‐access areas.

Hyperspectral images 3D reconstruction based on structure-from-motion and multi-view stereo

Infrastructure diagnostics, energy efficiency analysis, and remote inspection can benefit from access to thermal 3D models of buildings and other installations. However, photogrammetrically reconstructing 3D models from thermal image data alone usually produces insufficient results due to a lack of texture, uniform regions, overexposure, and inadequate resolution. Our approach decomposes the problem into two parts: reconstruction of the 3D model and association of the thermal information with the 3D model. For reconstruction, we investigated the use of RGB imagery to photogrammetrically reconstruct the 3D model using COLMAP for traditional Structure-from-Motion (SfM), as well as the use of emerging neural reconstruction methods such as Neuralangelo. For subsequent thermal information transfer, we propose two fusion strategies: (1) Manual registration of image-pairs using an affine transform derived from user-selected correspondences to map thermal values to mesh vertex attributes; (2) Calibration and correction of lens distortion (radial/tangential, principal-point shift) and sensor misalignment to consistently project thermal imagery onto the reconstructed model. We use a real-world case study of a university service building at Jade Hochschule (Wilhelmshaven, Germany), with imagery captured by a DJI Mavic 3T drone, to validate the methodology, assessing the trade-offs in reconstruction fidelity, processing time, and thermal data accuracy. While our results show that vertex-based fusion is feasible, labor-intensive manual registration and information loss due to cropping and vertex sparsity limit accuracy and usability. Requiring only one-time parameter estimation, calibration and texture-based thermal projection yields more realistic, higher-fidelity thermal overlays, while enabling largely automated processing after the initial estimation. As a fast alternative for ad hoc thermal scene inspection, we evaluated Instant-NGP and 3D Gaussian splatting. While both can visualize thermal appearance, they suffer from reduced quality compared to our RGB-based versions and exhibit artifacts. Overall, our results indicate that an RGB-first reconstruction pipeline with camera-calibrated thermal texturing is the most viable path to usable thermal digital twins from drone data, while neural reconstructions may capture complex surfaces but remain computationally costly. Our modular, reproducible framework for thermal 3D digital twins is available as open source.

Non-perennial streams, characterized by intermittent or episodic flows, comprise over half of global river networks and play an essential role in several ecosystem functions. Accurate stream channel topography is critical for representing flow, hyporheic exchange, and nutrient transport. Although many studies have applied Unmanned Aerial Vehicle (UAV)-based Structure-from-Motion (SfM) to reconstruct river and terrain topography, they have focused more on larger rivers or steep terrain and often relied on RTK-GNSS and ground control points (GCPs), leaving the performance of low-cost workflows for small non-perennial streams lacking evaluations. This study quantitatively evaluates the accuracy of multiple cost-efficient approaches for reconstructing 3-dimensional (3D) stream riverbeds: (1) a UAV imagery-based SfM approach, machine learning-based 3D reconstruction model, (2) Visual Geometry Grounded Deep Structure from Motion (VGGSfM), and (3) Visual Geometry Grounded Transformer for long sequence of images (VGGT-Long), and (4) handheld smartphone LiDAR scanning. The accuracy of the reconstructed topography was assessed against field measurements from a tripod optical level and GCPs GPS positions. UAV-based SfM emerges as the most effective and accessible method for accurately mapping non-perennial streambeds. Its planimetric error is around 1 m, and the ground elevation error is around 0.04 m. Although machine-learning based reconstructions substantially reduce computation time, they do not achieve comparable accuracy. Their planimetric error is over 5 m, and the ground elevation error is above 0.18 m. Likewise, iPhone LiDAR is not suitable for long reaches because cumulative sensor drift degrades positional and vertical precision, compromising the final reconstruction. Propagating these geometric errors into hydraulic and biogeochemical calculations showed that SfM yields relatively modest uncertainty in inferred water depth, velocity, nitrate uptake velocity, and reaeration, whereas the other methods introduce substantially larger uncertainty. This work exemplifies the significant potential for UAV-based surveys in characterizing stream habitats and conditions and in supporting reliable estimates of hydrobiogeochemical processes.

Image-based surface reconstruction and characterization is crucial for missions to small celestial bodies, as it informs mission planning, navigation, and scientific analysis. However, current state-of-the-practice methods, such as stereophotoclinometry (SPC), rely heavily on human-in-the-loop verification and high-fidelity a priori information. This paper proposes Photoclinometry-from-Motion (PhoMo), a novel framework that incorporates photoclinometry techniques into a keypoint-based structure-from-motion (SfM) system to estimate the surface normal and albedo at detected landmarks to improve autonomous surface and shape characterization of small celestial bodies from in situ imagery. In contrast to SPC, we forego the expensive maplet estimation step and, instead, use dense keypoint measurements and correspondences from an autonomous keypoint detection and matching method based on deep learning. Moreover, we develop a factor graph-based approach allowing for simultaneous optimization of the spacecraft’s pose, landmark positions, sun-relative direction, and surface normals and albedos via fusion of sun vector measurements and image keypoint measurements. The proposed framework is validated on real imagery taken by the Dawn mission to the asteroid 4 Vesta and the minor planet 1 Ceres and compared against an SPC reconstruction, where we demonstrate superior rendering performance compared to an SPC solution and precise alignment to a stereophotogrammetry (SPG) solution without relying on any a priori camera pose and topography information or humans-in-the-loop.

The integration of digital recording technologies into archaeological practice has fundamentally transformed how excavation processes are documented, analysed, and communicated.In recent years, three-dimensional recording techniques have increasingly enabled archaeologists to capture not only static archaeological remains but also the temporal dynamics of excavation itself.This topic formed a central theme of Session S24, "Digital Fieldwork Documentation in Archaeology: Innovations, Challenges and Standards," at the Computer Applications and Quantitative Methods in Archaeology (CAA) 2025 conference, where the work of Grigoriadi and Papaioannou was originally presented.Their study exemplifies how emerging digital documentation techniques can enhance both archaeological recording and the communication of excavation processes to wider audiences.The paper by Grigoriadi & Papaioannou (2026) presents a workflow combining Structure-from-Motion (SfM)photogrammetry with interactive 3D visualization to document the micro-excavation of a pot burial from the Phaleron Delta Cemetery in Attica, Greece.Through sequential image acquisition and photogrammetric reconstruction, the authors generated a series of three-dimensional models capturing successive microexcavation stages.These models were subsequently integrated into an interactive application developed in the Unity game engine, allowing users to explore both the spatial configuration and temporal progression of the excavation through a dynamic interface.This approach highlights the potential of digital documentation not

Accurate three-dimensional (3D) reconstruction of pavement is essential for structural health monitoring (SHM), in transportation infrastructure. However, conventional photogrammetric-based workflows under dynamic traffic scenarios face significant challenges due to vehicle occlusions, limited viewpoints and trade-offs between efficiency and cost. This study proposes an occlusion-aware, full-view digital pavement reconstruction framework integrating unmanned aerial vehicle (UAV)-based photography, an adaptive multi-scale vehicle detection network (AMVD-Net) and occlusion-aware multi-view optimisation into a conventional structure-from-motion (SfM) workflow. Field experiments across diverse road types indicate that AMVD-Net achieves a detection accuracy of 94.25% at 115.45 FPS, effectively eliminating vehicle-induced occlusions. The occlusion-aware multi-view optimisation maintains average relative errors below 1% and improves processing efficiency by 40–60% compared to the conventional SfM workflow. This scalable and field-validated UAV-SfM framework enhances automation, efficiency, and accuracy of pavement digitalisation, providing a robust foundation for digital twin applications and life-cycle management in transportation health monitoring.

Augmented Reality (AR) applications cannot be used without effective and robust 3D reconstructions of the scenes to appropriately position virtual objects into the real world. However, the reconstruction of high-fidelity 3D using mobile and resource-constrained hardware remains a significant challenge due to memory, processing, sensor and battery life constraints. Conventional methodologies that are geometric, such as Structure from Motion (SfM), Multi-View Stereo (MVS), and feature-based Simultaneous Localization and Mapping (SLAM) have also proven useful in tracking a camera and sparse-to-dense mapping. However, these techniques tend to perform poorly in low-texture scenes, moving scenes, and complicated lighting situations. This paper will solve these shortcomings by presenting a lightweight hybrid 3D reconstruction system that combines traditional SLAM approaches with a small neural augmentation system. Within this framework, SLAM will be used to precisely estimate poses and map geometries and the neural component will be used to refine critical regions to improve the quality of the texture and fill smaller gaps in reconstructions without causing a lot of computational load. The system also takes advantage of the performance of the embedded systems and client-grade GPUs, as well as, uses the GPUs to attain nearly real-time performance through the use of GPU-based optimization, lightweight data structure, and adaptive processing scheme. The conducted experiments suggest that the suggested hybrid scheme notably enhances the accuracy of reconstructions and visual quality without compromising on the performance specifications that are vital in resource-constrained settings.

Three-dimensional (3D) reconstruction using images is one of the most significant topics in computer vision and photogrammetry, with wide-ranging applications in robotics, augmented reality, and mapping. This study investigates methods of 3D reconstruction using video (especially monocular video) data and focuses on techniques such as Structure from Motion (SfM), Multi-View Stereo (MVS), Visual Simultaneous Localization and Mapping (V-SLAM), and videogrammetry. Based on a statistical analysis of SCOPUS records, these methods collectively account for approximately 6863 journal publications up to the end of 2024. Among these, about 80 studies are analyzed in greater detail to identify trends and advancements in the field. The study also shows that the use of video data for real-time 3D reconstruction is commonly addressed through two main approaches: photogrammetry-based methods, which rely on precise geometric principles and offer high accuracy at the cost of greater computational demand; and V-SLAM methods, which emphasize real-time processing and provide higher speed. Furthermore, the application of IMU data and other indicators, such as color quality and keypoint detection, for selecting suitable frames for 3D reconstruction is investigated. Overall, this study compiles and categorizes video-based reconstruction methods, emphasizing the critical step of keyframe extraction. By summarizing and illustrating the general approaches, the study aims to clarify and facilitate the entry path for researchers interested in this area. Finally, the paper offers targeted recommendations for improving keyframe extraction methods to enhance the accuracy and efficiency of real-time video-based 3D reconstruction, while also outlining future research directions in addressing challenges like dynamic scenes, reducing computational costs, and integrating advanced learning-based techniques.

Application of Structure-from-Motion (SfM) and Multi-View Stereo (MVS) to archival aerial photographs enables quantitative re-examination of terrain and surface changes caused by past disasters. This study reassessed three types of historical disasters－volcanic eruptions, landslides, and windthrow events－through differential analysis of pre- and post-event digital surface models (DSMs) reconstructed from the imagery. The SfM/MVS-derived DSMs captured volcanic terrain changes such as scoria-cone formation and lava-flow emplacement, and quantitatively delineated areas affected by landslides and windthrow.

Service crops are grown to provide ecosystem services in viticulture, but their adoption remains limited due to their competition with grapevine for soil resources. To identify trade-offs between services, the effect of service crops management strategies on grapevine performances still need further research. This dataset presents data from two experiments conducted to study the effect of service crops management on soil resources and grapevine performances. The inter-row vegetation was sampled in two Mediterranean vineyards using quadrats for biomass estimation. In addition, an unmanned aerial vehicle (UAV) was regularly flown over the vineyards for a period spanning more than four years in total over the two vineyards. The dataset presented here includes both raw data acquired during fieldwork and processed data derived from this raw inputs. The raw data consists of image series captured by two UAVs during each flight campaign, including RGB and multispectral imagery. Images were acquired between 2021–06–10 and 2022–07–29 for the first vineyard, and between 2023–06–08 and 2025–03–12 for the second vineyard. Based on these raw data, the processed data comprises spatial vectors, raster layers, and dense point clouds generated from UAV images using a Structure from Motion (SfM) photogrammetry workflow, at a 5 cm spatial resolution. The raster layers and dense point clouds provide specific information on vineyard characteristics for each UAV flight date, including elevation, vegetation indices, visible and near-infrared reflectance, and canopy height. In addition, the processed data include measurements of vegetation dry biomass, as well as separate measurements of dry biomass and leaf area measured for selected service crops species. This dataset can be reused for the calibration and/or evaluation of classification algorithms aimed at discriminating vines from the inter-row vegetation, or as part of a larger dataset to explore relationships between remotely-sensed vegetation indices and field-measured vegetation biomass or surface.

Abstract A fundamental understanding of landslide evolution requires characterizing how deformation localizes within the sliding mass, as these non-homogeneous zones provide crucial insights into how destabilization initiates, failure surfaces develop, and the overall kinematic behavior evolves. While traditional analysis often assumes uniform movement, this study presents a methodology to quantify intricate patterns of surface deformation at a fine scale, allowing for the direct analysis of localization behavior. By applying strain tensor analysis to high-resolution displacement fields derived from multi-temporal Uncrewed Aerial Vehicle-Light Detection and Ranging (UAV-lidar) and Structure from Motion (SfM) surveys, we compute the divergence, gradient, and curl fields for two distinct landslides: one translational and one rotational. This approach quantifies volumetric changes, translational strain, and rotational components, revealing unique kinematic signatures for each landslide type. The translational slide is characterized by alternating expansion-contraction patterns along its dip-line, whereas the rotational slide exhibits clear, separate bands of head subsidence and toe expansion, coupled with non-uniform rotation along the strike. This detailed characterization of strain localization provides direct observational evidence of the fundamental mechanisms governing landslide behavior. It offers a more nuanced, mechanistic understanding that advances the interpretation of slope instability, providing a stronger physical basis for hazard assessment and risk management.

Coral reef ecosystems are increasingly threatened by climate change and human activities, and conservation organizations worldwide require efficient, cost-effective tools to document habitat conditions, assess artificial reef growth, and monitor corallivore dynamics. Traditional approaches, such as transect surveys with visual inspection, often suffer from low efficiency, limited coverage, and insufficient spatial information. To address these limitations, we developed a low-cost and versatile underwater camera system with three operating modes: surface, horizontal underwater, and curved underwater. The system records continuous video through three synchronized cameras, which are processed using Structure-from-Motion (SfM) photogrammetry to generate 3D reconstructions for habitat and colony-scale monitoring. A point cloud-based semantic segmentation model (KPConv) was then applied to segment the reconstructed point clouds into 'Drupella snails', 'Compromised coral', and 'Others', thereby indicating coral conditions affected by corallivore predation. Field surveys conducted in Koh Tao, Thailand, demonstrated the feasibility of this integrated hardware-software approach. The camera system achieved higher point densities and lower costs compared to conventional methods, while remaining lightweight and cost-effective for community-based conservation. The KPConv model yielded an overall accuracy of 0.916 and a mean Intersection over Union of 0.601, enabling the identification of compromised corals and cryptic Drupella snails. Quantitative assessments based on segmented outputs further revealed coral conditions and provided early warnings of potential Drupella snails outbreaks. In summary, this novel system integrates affordable image collection hardware with 3D reconstruction and semantic segmentation, improving survey efficiency and providing spatially explicit insights for coral reef monitoring, particularly for small-scale corallivore dynamics.

We examined perceived reversal rates and biases as observers viewed four ambiguous, motion-defined depth asymmetric point-light stimuli: a biological motion stimulus in the form of a spinning point-light walker (PLW), a rigidly spinning human figure, a spinning half-cylinder, and a wobbling slanted cylinder. The last three are rigid structure-from-motion (SFM) stimuli. We analysed angular reversal distributions to assess perceptual biases: facing-the-viewer (FTV), convexity, and depth-symmetry biases. The PLW showed the highest reversal rate and a strong FTV bias, though responses were bimodal, some observers experienced reversals every half-turn, others rarely. The rigid human figure showed a weak FTV bias. The spinning half-cylinder resulted in an initial convexity bias, but the occurrence of reversals following the initial percept instead revealed a novel "edge-in-front" bias. The wobbling cylinder showed no angular bias, and had the fewest reversals, likely due to persistent depth asymmetry throughout its motion, and/or that the wobbling prevent adaptation-recovery cycles of neural populations tuned to opposite spinning directions. Correlation analyses revealed shared mechanisms among spinning stimuli, but not with the wobbling cylinder. These findings highlight how shape and motion jointly influence perceptual reversals, refining models of bistable perception and individual variability.

ABSTRACT This study aims to evaluate the dynamics of a semi-hanging beach–dune system in an open sandy beach during winter and summer 2021–2022, using UAV-based Structure-from-Motion (SfM) and Multi-View Stereo (MVS) surveys. Morphometric and morphodynamic parameters were extracted from DEM-derived cross-shore profiles and 3D clouds were compared. Oceanographic forcings were analysed to identify storm events and dune impact hours, while sediment volume balance was quantified. Results highlighted the critical role of winter storms (May–September), where five extreme events triggered foredune erosion of 8600 m3, with an average dune toe retreat of 3.5 m. Sediments were redistributed on the beach, accumulating 4100 m3, favouring a storm berms and cusp formations, completely eroded in summer. Over the annual period, the foredune was unable to recover from the storm-induced sediment deficit. Post-storm alongshore variability was mainly controlled by antecedent beach and dune slopes, dune toe height, and the cumulative duration of impact hours. Overall, the results indicate a clear sedimentary imbalance in the system, characterized by high vulnerability to severe events and low resilience. This study contributes new insight into the morphodynamic behaviour and resilience of semi-hanging beach–dune on the Buenos Aires coast and demonstrates the effectiveness of UAV-based SfM-MVS monitoring in capturing coastal morphodynamics.

This study presents a real-time framework for generating two-dimensional (2D) orthomosaic maps directly from UAV video. The method targets operational scenarios in which a continuously updated 2D overview is required during flight or immediately after landing, without relying on time-consuming offline photogrammetry workflows such as structure-from-motion (SfM) and multi-view stereo (MVS). The proposed procedure incrementally registers sparsely sampled video frames on standard CPU hardware using classical feature-based image registration. Each selected frame is converted to grayscale and processed under a fixed keypoint budget to maintain predictable runtime. Tentative correspondences are obtained through descriptor matching with ratio-test filtering, and outliers are removed using random sample consensus (RANSAC) to ensure geometric consistency. Inter-frame motion is modeled by a planar homography, enabling the mapping process to jointly account for rotation, scale variation, skew, and translation that commonly occur in UAV video due to yaw maneuvers, mild altitude variation, and platform motion. Sequential homographies are accumulated to warp incoming frames into a global mosaic canvas, which is updated incrementally using lightweight blending suitable for real-time visualization. Experimental results on three UAV video sequences with different durations, flight patterns, and scene targets report representative orthomosaic-style outputs and per-step CPU runtime statistics (mean, 95th percentile, and maximum), illustrating typical operating behavior under the tested settings. The framework produces visually coherent orthomosaic-style maps in real time for approximately planar scenes with sufficient overlap and texture, while clarifying practical failure modes under weak texture, motion blur, and strong parallax. Limitations include potential drift over long sequences and the absence of ground-truth references for absolute registration-error evaluation.