Conference Agenda

Cultivated land conversion to built-up area is a core form of farmland non-agriculturalization and a major threat to farmland protection in China. Current remote sensing methods for detecting such changes face two limitations: insufficient integration of domain prior knowledge and the inability of purely data-driven models to achieve both high Precision and Recall. To address these issues, this study proposes a knowledge graph-enhanced change detection method. A multi-scale knowledge analysis framework incorporating feature, scene, and business knowledge layers is constructed to systematically integrate multi-source geographic information into structured semantic representations. A knowledge fusion residual network, KG-MS-ResNet, is designed based on ResNet-18 with modifications to the first convolutional layer for bi-temporal image inputs. TransE embeds geographic indicator knowledge into multi-scale semantic vectors, while a semantic–feature dual-path fusion strategy and a knowledge-guided attention mechanism enable deep coupling between image features and domain knowledge. Experiments in Pei County, Jiangsu Province, show that the proposed method outperforms baseline ResNet across all metrics, with Recall increasing by 4.84 percentage points and F1-score by 0.0752. The results demonstrate that integrating domain knowledge graphs with deep learning significantly improves detection performance, offering a semantically interpretable solution for monitoring cultivated land non-agriculturalization and advancing the integration of knowledge-driven and data-driven approaches in intelligent remote sensing interpretation.

Road change detection is essential for maintaining up-to-date digital maps; however, conventional update processes rely heavily on the manual interpretation of aerial imagery, leading to high labor costs and inconsistent outcomes. To address these limitations, this study proposes an automated road change detection method that integrates aerial orthophoto-based segmentation with geometric boundary deviation analysis. Road areas are first extracted from high-resolution aerial orthophotos using SegFormer, a Transformer based semantic segmentation model. The segmentation results are then converted into vector polygons for geometric analysis. Structural changes, such as newly constructed or removed roads, are detected through a difference-based comparison with historical digital maps. Simultaneously, shape changes are quantitatively analyzed by measuring geometric deviations between road boundaries.

Specifically, vertex-wise distances between corresponding boundaries are computed, and the overall deformation is evaluated using Root Mean Square Error (RMSE), incorporating Z-score-based outlier removal to ensure robustness against noise. Experimental results demonstrate that the proposed method effectively detects both structural changes and subtle geometric variations, including road expansions and boundary shifts. Furthermore, the method enables clear object-level classification of change types, providing a practical and efficient framework for digital map updating workflows.

Accurate building change detection depends on how well building boundaries are delineated, as distortions and merging errors hinder reliable correspondence. In dense urban areas, deep learning models frequently merge adjacent buildings—especially within narrow gaps—producing structural inconsistencies that lead to change detection errors. We propose a post-processing method integrating Local Rank-Based Prior Calibration, which reinterprets Softmax probabilities as percentile-based local ranks, with Graph-Cut refinement for structural correction. The refined mask is matched with historical building data to classify four change types. Experiments using aerial imagery from Seoul show that the method reduces structural errors, lowering under-segmentation from 51.64% to 22.02% and improving IoU from 0.748 to 0.759. In change detection, it increases the mean F1-score from 0.522 to 0.608 and improves all classes, including new construction, whose F1-score rises from 0.269 to 0.707. Ablation studies confirm that calibration and graph-based refinement both contribute to the improvements. These results show that stabilizing segmentation outputs enhances the reliability of building-level change detection in dense urban environments.

OpenStreetMap (OSM) is a representative open geospatial platform that provides free access to major spatial objects, including buildings worldwide, constructed through crowdsourcing-based manual digitization. However, subjective differences among contributors and the absence of unified quality control standards have led to the accumulation of positional offsets and boundary shape errors in building polygons. To address this issue, studies using deep learning-based semantic segmentation for OSM quality improvement have been conducted. Nevertheless, Transformer-based segmentation models exhibit an under-segmentation tendency that merges adjacent buildings into a single object, along with limitations in precise boundary delineation. To overcome these challenges, this study proposes a two-stage framework that integrates SegFormer, which excels in global context recognition, with SAM 2, which is capable of precise boundary segmentation. In the first stage, SegFormer semantically segments building regions from a true orthoimage, and in the second stage, SAM 2 infers object-level precise boundaries using the bounding boxes of OSM polygons as box prompts. The two results are combined into a prior probability map, enabling uncertain boundary regions to be re-evaluated in an unsupervised manner. In experiments conducted over the Suseo-dong area in Gangnam-gu, Seoul, the proposed method achieved a BIoU of 70.40%, an improvement of 23.85 percentage points over OSM building data, with consistent performance gains across all evaluation metrics. This framework offers scalability applicable to any region worldwide without additional label construction, provided that high-resolution true orthoimagery and OSM data are available.

Accurate building footprint extraction is critical for applications ranging from population estimation to disaster management. Although optical imagery provides detailed spectral information, it often struggles with shadows, occlusions, and background clutter in dense urban environments. Lidar data, by contrast, offer precise elevation and structural attributes but face challenges such as variable point density and noise. This study integrates multispectral imagery from the U.S. Department of Agriculture (USDA) National Agriculture Imagery Program (NAIP) with lidar-derived feature height and intensity from the U.S. Geological Survey (USGS) 3D Elevation Program (3DEP) to improve footprint extraction using a U-Net–based deep learning model. A six-band input stack (RGB, near-infrared, height, intensity) was developed, normalized, and tiled for training and evaluation against Microsoft Global Building Footprints (GBF). Results from the Houston, TX test site show that the six-band model achieved a precision of 0.86, recall of 0.88, F1 score of 0.87, and Intersection-over-Union (IoU) of 0.76, consistently outperforming four-band baselines by reducing false positives while maintaining sensitivity. Predictions on withheld Houston tiles confirmed strong within-region generalization, yielded a precision of 0.78, recall of 0.81, F1 score of 0.79, and IoU of 0.66. Qualitative analysis further revealed limitations stemming from both training label quality and vegetation–building confusion. These findings demonstrate the complementary value of integrating spectral and structural information for robust building footprint extraction and how domain adaptation strategies can be used to enhance cross-regional transferability.

This contribution presents a benchmarking study of a lightweight deep learning model for automatic pothole detection in asphalt pavements. Accurate and cost-effective identification of surface distresses is essential for road safety and for prioritising maintenance, especially in cities where traditional visual surveys are still predominant. We adapt and train a compact YOLO-based object detection architecture on a dataset of annotated street-level images, covering different lighting conditions, pavement textures and distress severities. The study evaluates how input resolution, confidence thresholds and data augmentation strategies affect detection performance and inference speed, and compares the lightweight model with heavier state-of-the-art detectors. Results indicate that it is possible to obtain competitive accuracy while maintaining real-time processing capabilities on modest hardware, which is crucial for deployment in mobile inspection platforms such as smartphones, dashcams or low-cost onboard units. The paper discusses opportunities and limitations of integrating deep learning into pavement management systems and outlines perspectives for extending the approach to other types of defects and to larger road networks.

This study presents a deep learning framework integrating Sentinel‑2, Sentinel‑1, and GEDI LiDAR to map Ethiopia’s canopy height at 10‑m resolution from 2019–2024. A shift‑aware loss function was employed to correct geolocation errors inherent in GEDI L2A footprints, and height‑weighted penalties addressed systematic underestimation in tall forests. Results show a national net gain of 23,537 km² in tree cover >8 m, reversing long‑standing deforestation trends. Gains concentrated in low‑to‑mid canopy strata (<20 m), strongly associated with major restoration interventions including the Green Legacy Initiative (GLI), REDD+, and the Sustainable Land Management Program (SLMP). Losses persist in western and southeastern highlands, driven by agricultural expansion, wildfires, infrastructure development, and large‑scale agricultural investments. This work demonstrates the operational value of multi‑sensor deep learning for near‑real‑time monitoring of restoration outcomes in data‑scarce regions.

Accurate tree-level analysis in forests via LiDAR scanning is essential for biomass estimation, canopy structure assessment, and carbon monitoring, yet remains constrained by the scarcity of large-scale annotated LiDAR datasets and the high cost of manual annotation. To address this, we present a novel approach that integrates 3D tree models with UAV-borne LiDAR simulation to generate synthetic forest point clouds with comprehensive annotations. Our approach generates diverse woodland, open, and closed forest structures, producing Synthetic Forest, a benchmark datasets of three 1 ha scenes containing 38–47 million points each, with densities of 3300–3860 points/m² and average spacing of 2 cm. Each scene contains between 70 and 216 individual trees, along with understory vegetation, deadwood, stumps, rocks, and bushes, all automatically annotated with semantic classification IDs, instance IDs, and tree IDs for volume estimation. The proposed pipeline provides automated, error-free ground truth for leaf-wood classification, instance segmentation, and wood volume estimation. We provide a guideline for generating forest plots and utilizing the datasets for diverse forestry tasks. By eliminating the need for costly field data collection, our pipeline offers scalable, customizable synthetic datasets that accelerate forest inventory. The Synthetic Forest dataset is publicly released via Zenodo (DOI: 10.5281/zenodo.17568131), enabling reproducible research and supporting further developments in forest monitoring and management.

Accurate mapping of tree species is essential for forest monitoring, biodiversity assessment, and ecological applications. Very high-resolution UAV imagery provides detailed structural and spectral information, but species-level segmentation remains challenging due to limited annotated data, complex crown geometries, and strong visual similarity among taxa. Recent Remote Sensing Foundation Models (RSFMs) offer new possibilities by providing transferable representations learned from large, multimodal geospatial datasets.

This contribution introduces a two-phase framework that combines foundation model initialization with multi-temporal UAV imagery to enhance fine-grained forest segmentation. In Phase 1, a DeepLabv3+ network is initialized using FoMo-Net, a ViT-based RSFM pre-trained on the multi-scale FoMo-Bench benchmark. This initialization enables strong generalization from heterogeneous global forest datasets to very high-resolution UAV scenes. In Phase 2, phenological information is integrated by fusing May and September UAV acquisitions through temporal difference composites and pseudo-label refinement, allowing the model to resolve species-specific seasonal patterns.

Experiments on the Québec Trees Dataset, covering 14 species at 0.02 m GSD, demonstrate substantial performance gains. Foundation model initialization improves overall accuracy from 52.79% to 71.21%, while incorporating multi-temporal cues further increases accuracy to 78.21%. The results highlight the complementary roles of structural priors learned by RSFMs and phenological information captured by UAV time series for detailed forest species mapping.

Accurate mapping of forest types and vegetation characteristics is essential for monitoring biodiversity and forest dynamics. Traditional Deep Learning (DL) models trained on Sentinel-2 time series achieve high performance, but require extensive preprocessing and sensor-related fine-tuning. In this study, we evaluate the recently introduced AlphaEarth Foundations (AEF) embeddings, which is global, multi-modal feature representation of the earths surface, for forest mapping in Italy. We compare a) a Multi-Layer Perceptron trained on AEF, b) a Time-Series Transformer trained on Sentinel-2 annual time series and CHELSA climate data, and c) a Cross-Attention fusion model combining both feature sets. Using 5-fold cross-validation in a regression and a classifaction task on two datasets (evergreen broad-leaved tree cover ETC, forest vegetation type FVT) we find that the combined model consistently outperforms the single-source approaches (RMSE = 0.161, Accuracy = 0.757). AEF-based models achieve comparable accuracy to the Sentinel-2-based model while reducing extensive time series preprocessing and training time by an order of magnitude. Feature attribution using integrated gradients reveals that AEF provides stable baseline representations, while Sentinel-2 inputs add phenology-related detail. The results show, that integrating generalized embeddings with specialized spectral-temporal features improves predictive performance for forest mapping.

In Japan, many urban park trees are becoming large and aged, increasing the risk of structural failures caused by extreme weather events and biological deterioration. Effective management therefore requires reliable risk assessment, for which accurate tree species identification is one of the fundamental prerequisites. However, species identification still depends heavily on visual assessment by skilled professionals, posing challenges in efficiency and objectivity. This problem is particularly significant for broad-leaved trees, which exhibit high species diversity and morphological variability. In addition, labor shortages have intensified the demand for automated and reliable classification techniques. This study proposes a high-accuracy classification method for broad-leaved tree species using ground-level images captured with a commercially available RGB camera and deep learning. The proposed method extracts small local patches that capture species-specific visual features, such as leaf shape and bark texture, commonly used by professional arborists for species identification. These local features are evaluated individually using deep learning models, allowing fine-scale visual characteristics to be effectively utilized for classification. To address variability in outdoor imaging conditions, including illumination changes, shadows cast by branches and leaves, and moss attachment, multiple patches are classified independently and the results are integrated through majority voting, improving classification robustness. Experiments were conducted on seven tree species commonly found in Japanese urban parks: cherry, ginkgo, zelkova, konara oak, sawtooth oak, plane tree, and flowering dogwood. The results demonstrate that the proposed method achieves a maximum classification accuracy of approximately 95% under real-world conditions, demonstrating its effectiveness for practical urban tree management.

This study compares visualization types (3D, 2D, Oblique 3D, and Combined 2D+3D, coupled in a pairwise fashion for different tasks and scenarios), investigates their influence on decision-making across selected urban planning tasks (Site selection, Scenario Selection), and Distance Estimation as a baseline task that we assumed is relevant in both. Our goal was to inform the participatory urban planning process. In a controlled user study with 40 participants, we evaluated whether visualization type affects decision outcomes and distance estimation, complemented by participants’ visualization preferences before and after the experiment. The results confirm the previously well documented evidence that participants are considerably more successful with distance estimation with 2D visualizations, and their decisions vary depending on the examined visualization and task types. We observe that different formats support different task requirements, i.e., each visualization type exhibits distinct strengths depending on the task. These findings indicate that visualization choice in urban planning should be adapted to the task and context rather than treated as an interchangeable artifact.

With the rising number and severity of WUI wildfire episodes and the necessity to improve community preparedness, planning strategies have to be devised that integrate foresight into wildfires, with active community participation. This paper presents an intelligent collaborative environment that seeks to engage citizens, planners, and emergency services in co-creation of fire-resilient strategies through Agentic AI-driven wildfire simulations and mixed-reality visualization. A serious game environment is designed for hands-on exploration of alternative wildfire spread scenarios and community-scale prevention practices such as prescribed burning, fuel treatment through vegetation control, and structural hardening measures. The objective is to promote public awareness and adaptive behavior as well as provide science-based operational decision support to emergency responders in evaluating tactical options specific to terrain, infrastructure, and fuel conditions of the locale. The system operates on a Large Language Model (LLM)–powered agentic AI architecture designed to automate and orchestrate 2D and 3D wildfire simulations, providing guidance that supports users from diverse technical backgrounds. To give the results of the simulations, 3D web visualization and immersive holographic display were used to enable cycles of iterated explorations into fire spread in dense urban settings. With AI-assisted wildfire intelligence, this particular flow works through a set of intuitive interaction mechanics so that communities can evaluate risk levels, weigh alternatives for mitigation, and better prepare for an actual fire event.

Accurate spatial alignment is a key requirement for situated Augmented Reality (AR) in urban planning, where citizens and planners can visualize proposed designs in real outdoor environments. However, existing AR localization approaches often rely on smartphone GNSS, vendor-specific cloud anchors, or cloud-based visual positioning, which introduce accuracy limitations, privacy concerns, or dependencies that restrict their use in participatory planning workflows. This paper presents a privacy-aware on-device localization pipeline for outdoor urban planning scenarios. The approach aligns LiDAR scans captured on smartphones with pre-scanned reference point cloud tiles to enable stable and accurate placement of urban planning models. Approximate GNSS is used only to retrieve a relevant reference tile, while all preprocessing and registration steps are performed locally on the device. The pipeline combines voxel downsampling, local geometric descriptors, and global registration to estimate alignment without relying on GNSS for pose estimation or on cloud-based visual localization services. A mobile demonstrator was developed to support situated AR in urban planning scenarios, allowing users to explore design proposals directly in context. Initial validation under controlled conditions showed that the system can recover translations and rotations with errors on the order of a few centimeters, while processing times remained suitable for mobile use. The approach was also deployed in an urban planning case study and enabled stable outdoor visualization of planning elements on-site.

We present a new method for assessing visual walkability using 360° videos and an eye-tracking in Cinematic Virtual Reality (CVR). Visual walkability refers to the walkability perceived by pedestrians through visual stimuli in the urban environment. Our method uses semantic segmentation, viewport exposure, gaze measures, and a custom walkability questionnaire, enabling comparison between scene content, participant's viewport, and their gaze focus. The 10 videos used, including 2 calibration videos, exhibit distinct semantic characteristics, validated by segmentation analysis. Analysis of the 35 participants’ responses shows that walkability ratings at the video level correlate with several environmental parameters (e.g., road, sidewalk, sky) consistent with previous studies. However, these parameters do not have a similar influence in gaze-based visual attention analysis within the CVR setting, suggesting that CVR attention would requiere further work. Furthermore, our results suggest that unexpected semantic classes may also play a role in perceived walkability and should be considered exploratory pending further validation. This paves the way for further research on using CVR as an assessment tool for visual walkability and for developing methodological guidance on which visual cues are robust across measures (content/viewport).

Oceanographic data and their related simulation have a key role in addressing EU and UN societal challenges in marine environments. Visualising marine data is challenging for different visual-communication intents and audiences, despite existing guidelines on the subject. A main visual-design limitation for existing techniques is the co-visualization of multiple hydrodynamic field attributes in an accessible, comprehensible and engaging manner. This paper addresses this limitation in two ways: first, existing techniques for cartographic-oriented design of waterlines are adopted and extended towards multivariate hydrodynamic field datasets. Secondly, experimental results on the intermixing different visual-channel mapping of hydrodynamic attribute data are presented in a case study on ocean-flow patterns around the Hebrides island chain (UK). The results demonstrate a simultaneous co-visualization of up to five unique, independent scalar attributes in a comprehensible manner while preserving the geographic context. Moreover, best-practice guidelines are stated in conclusion of the experimental case study to help oceanographic practitioners adopt the presented technology in their professional workflows.

Artificial light at night (ALAN) degrades nocturnal ecosystems and complicates astronomical observation. Although all-sky imaging and GIS-based light-pollution mapping are well established in the analysis of light pollution, identifying local contributors to ALAN still requires time-consuming cross-comparisons, done in separate views, making light halo--source attribution slow and manual. We present an interactive system that addresses this gap by co-registering Sky Quality Camera all-sky imagery and OSM-derived candidate emitters (e.g., settlements, roads, aerodromes, industrial sites) in one observer-centered scene. The viewer is placed at the locations of the captured all-sky images in 3D digital terrain model-based scenes, realistically illuminated by the sky under selected conditions for an immersive view of nighttime scenarios. OpenStreetMap features are projected onto a surrounding sphere via inverse stereographic projection, with point markers and horizontal-extent indicators to support rapid visual matching between observed halos and plausible sources. Users can switch scenes and processed sky images, adjust projection parameters, and inspect scenes in VR or in an additional cylindrical projection for a panoramic desktop view. A companion web tool configures location classes and display ranges. The presented system primarily targets exploratory analysis, with its main contribution being the novel co-visualization of light sources and light halos; expert interviews positively validated this analytical focus. As a secondary outcome, the system's immersive first-person representation may also enrich educational communication and outreach on ALAN impacts.

Accurate and intuitive visualization of urban development projects is a persistent challenge in spatial planning and public participation. Recent advances in Extended Reality (XR) offer new opportunities to integrate geospatial data directly within the user’s real environment. This paper introduces GeoWebXR, an extension of the WebXR API designed to provide absolute georeferencing of the XR reference space via a standardized geopose. We present an outdoor proof-of-concept implementation that integrates a dual-antenna RTK GNSS receiver mounted on an XR headset. High-precision GNSS measurements are fused with the device’s local pose estimates to compute a consistent and accurate geopose, enabling decimeter-level alignment between virtual and physical environments.

Leveraging GeoWebXR, WebGL applications can render georeferenced 3D content in situ through a web browser. We demonstrate this capability using the iTowns geospatial visualization framework to deliver an XR experience for urban planning. The system supports both 1:1-scale in-situ visualization and reduced-scale overview modes, enabling seamless multiscale exploration of planning scenarios. To mitigate cognitive overload in dense urban contexts, we implement and evaluate several visualization and interaction strategies. We assess the usability and spatial appropriation enabled by our system, and discuss how it may support both expert analysis and citizen participation in urban planning processes.