Wouter A. J. Van den Broeck

Mr. Wouter A. J. Van den Broeck

orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
22 November 2022 in Remote Sensing
Verified
Journal Article
Open access
PDF file
WBWouter A. J. Van den Broeck
Wouter A. J. Van den Broeck
Toon GoedeméToon Goedemé
Toon Goedemé
MLMaarten Loopmans
Maarten Loopmans
Historical land cover (LC) maps are an essential instrument for studying long-term spatio-temporal changes of the landscape. However, manual labelling on low-quality monochromatic historical orthophotos for semantic segmentation (pixel-level classification) is particularly challenging and time consuming. Therefore, this paper proposes a methodology for the automated extraction of very-high-resolution (VHR) multi-class LC maps from historical orthophotos under the absence of target-specific ground truth annotations. The methodology builds on recent evolutions in deep learning, leveraging domain adaptation and transfer learning. First, an unpaired image-to-image (I2I) translation between a source domain (recent RGB image of high quality, annotations available) and the target domain (historical monochromatic image of low quality, no annotations available) is learned using a conditional generative adversarial network (GAN). Second, a state-of-the-art fully convolutional network (FCN) for semantic segmentation is pre-trained on a large annotated RGB earth observation (EO) dataset that is converted to the target domain using the I2I function. Third, the FCN is fine-tuned using self-annotated data on a recent RGB orthophoto of the study area under consideration, after conversion using again the I2I function. The methodology is tested on a new custom dataset: the ‘Sagalassos historical land cover dataset’, which consists of three historical monochromatic orthophotos (1971, 1981, 1992) and one recent RGB orthophoto (2015) of VHR (0.3–0.84 m GSD) all capturing the same greater area around Sagalassos archaeological site (Turkey), and corresponding manually created annotations (2.7 km² per orthophoto) distinguishing 14 different LC classes. Furthermore, a comprehensive overview of open-source annotated EO datasets for multiclass semantic segmentation is provided, based on which an appropriate pretraining dataset can be selected. Results indicate that the proposed methodology is effective, increasing the mean intersection over union by 27.2% when using domain adaptation, and by 13.0% when using domain pretraining, and that transferring weights from a model pretrained on a dataset closer to the target domain is preferred.

21 September 2022 in Remote Sensing
Verified
Journal Article
Open access
PDF file
WBWouter A. J. Van den Broeck
Wouter A. J. Van den Broeck
Toon GoedeméToon Goedemé
Toon Goedemé
The roofscape plays a vital role in the support of sustainable urban planning and development. However, availability of detailed and up-to-date information on the level of individual roof-part topology remains a bottleneck for reliable assessment of its present status and future potential. Motivated by the need for automation, the current state-of-the-art focuses on applying deep learning techniques for roof-plane segmentation from light-detection-and-ranging (LiDAR) point clouds, but fails to deliver on criteria such as scalability, spatial predictive continuity, and vectorization for use in geographic information systems (GISs). Therefore, this paper proposes a fully automated end-to-end workflow capable of extracting large-scale continuous polygon maps of roof-part instances from ultra-high-resolution (UHR) aerial imagery. In summary, the workflow consists of three main steps: (1) use a multitask fully convolutional network (FCN) to infer semantic roof-part edges and objects, (2) extract distinct closed shapes given the edges and objects, and (3) vectorize to obtain roof-part polygons. The methodology is trained and tested on a challenging dataset comprising of UHR aerial RGB orthoimagery (0.03 m GSD) and LiDAR-derived digital elevation models (DEMs) (0.25 m GSD) of three Belgian urban areas (including the famous touristic city of Bruges). We argue that UHR optical imagery may provide a competing alternative for this task over classically used LiDAR data, and investigate the added value of combining these two data sources. Further, we conduct an ablation study to optimize various components of the workflow, reaching a final panoptic quality of 54.8% (segmentation quality = 87.7%, recognition quality = 62.6%). In combination with human validation, our methodology can provide automated support for the efficient and detailed mapping of roofscapes.

16 June 2025 in Methods in Ecology and Evolution
Verified
Journal Article
Open access
PMPietari Mönkkönen
Pietari Mönkkönen
WBWouter A. J. Van den Broeck
Wouter A. J. Van den Broeck
SASimo Ali-Löytty
Simo Ali-Löytty
Kim CaldersKim Calders
Kim Calders
PRPasi Raumonen
Pasi Raumonen
Leaves are crucial for tree growth, survival and interaction with the environment. The spatial arrangement of leaves within the tree canopy significantly influences tree‐environment interactions, such as sunlight capture, scattering and shading, as well as rainfall interception. The impact of leaf configurations on ecological processes can be studied using virtual 3D foliage models. These models are valuable for ecological simulations and modelling applications, as well as for the development and validation of remote sensing methods. This virtual approach enables the study of leaf attributes and interactions in a controlled, reproducible and cost‐effective manner. The study introduces LeafGen, a new MATLAB‐based foliage generation method to generate explicit leaves on 3D tree models. The leaf cover is generated based on user‐defined leaf area density, orientation, and size distributions with flexible and configurable distribution types and parameters. The foliage can be generated either on cylinder‐based 3D tree models (quantitative structure models) or point clouds of trees. The distributions are defined on structural variables of the tree models, making the definition straightforward and simple but allowing diverse configurations. Species‐specific traits can also be incorporated with specific leaf geometries and phyllotaxis patterns. The leaves are generated without intersections to ensure realistic foliage. LeafGen provides an easily accessible and low‐entry tool for ecological research, enabling precise control over leaf attributes and their distribution. The method is open source and has a publicly hosted repository on GitHub along with guidance on implementing different scenarios. This method enhances the reproducibility and cost‐effectiveness of studies on leaf–environment interactions and can be applied to various applications, including forest remote sensing, light transfer modelling and tree growth simulations.

14 September 2026 in Communications Earth & Environment
Verified
Journal Article
Open access
PDF file
LMLion R. Martius
Lion R. Martius
PSPablo Sanchez-Martinez
Pablo Sanchez-Martinez
VNVanessa Negrão-Rodrigues
Vanessa Negrão-Rodrigues
PBPaulo Bittencourt
Paulo Bittencourt
WCWout Cherlet
Wout Cherlet
PMPatrick Meir
Patrick Meir
The rising frequency and intensity of extreme droughts across the Amazon raises uncertainty about the capacity of tropical forests to adjust to future climate. Here, we combined the severe 2023 El Niño drought with a multi-decadal throughfall-exclusion experiment to compare eco-physiological resilience between pre-exposed drought-primed trees (PT) and untreated controls (CT). During El Niño, PTs reduced transpiration by 25%, while CTs declined by up to 54% relative to pre-drought levels. Post-drought, PTs fully recovered transpiration, whereas CTs recovered only to 77.5%, showing legacy effects and pronounced leaf area loss. These differences were reinforced by long-term structural changes (i.e., mortality) in the TFE, which reduced water competition during extreme drought. Tree structural data revealed a significantly smaller crown-to-stem size ratio in PTs, reducing sensitivity to evaporative demand and underpinning their enhanced drought resistance and resilience. These findings reveal that physiological resilience to extreme events can emerge from long-term drought exposure through structural adjustments at both ecosystem and individual tree scale.