Edward Verbree
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Edward Verbree has a total of 51 co-authors. Here are the top co-authors by number of shared publications:
- Sisi Zlatanova
University of New South Wales
4 shared publications - Martijn Meijers
Delft University of Technology
3 shared publications - Peter van Oosterom
Delft University of Technology
3 shared publications Peter van Oosterom
2 shared publicationsYingwen Yu
2 shared publicationsZhuoyue Wang
2 shared publications
Latest Publications
- Open Access
- 18 November 2026
- Published by Springer Nature in npj Heritage Science
LiDAR assisted 3D Gaussian splatting for surface deterioration mapping in semi-subterranean built heritage
- 18 November 2026
- Published by Springer Nature in npj Heritage Science
Open AccessLiDAR assisted 3D Gaussian splatting for surface deterioration mapping in semi-subterranean built heritage
- Open Access
- 30 October 2026
- Published by TU Delft OPEN Publishing
To Dredge or not to Dredge: Data-driven Feature Engineering of Side Channels
Side channels improve river resilience by increasing flow capacity and providing water buffer space. Keeping these side channels flowing properly is crucial for preventing fluvial flooding, especially given the more extreme weather caused by climate change. In this paper, we extract insights directly from measured geographic data in order to predict sedimentation behavior of side-channels within the Dutch Waal, Lek, and Nederrijn rivers. We investigated how features of a side channel can be identified and extracted to enable predictive maintenance. By reviewing existing literature, key side channel characteristics are identified and captured into machine-readable features using an automatic pipeline. Using these features, a proof-of-concept machine learning model is trained to predict sedimentation, achieving 84% accuracy with a 5cm error margin for a case study channel. The result is an automated node-based data processing pipeline using the FME software, alongside a geospatial database with 19 features for 26 side channels, and a machine learning model showcasing the predictive ability of the derived features.
- 30 October 2026
- Published by TU Delft OPEN Publishing
Open AccessTo Dredge or not to Dredge: Data-driven Feature Engineering of Side Channels
Side channels improve river resilience by increasing flow capacity and providing water buffer space. Keeping these side channels flowing properly is crucial for preventing fluvial flooding, especially given the more extreme weather caused by climate change. In this paper, we extract insights directly from measured geographic data in order to predict sedimentation behavior of side-channels within the Dutch Waal, Lek, and Nederrijn rivers. We investigated how features of a side channel can be identified and extracted to enable predictive maintenance. By reviewing existing literature, key side channel characteristics are identified and captured into machine-readable features using an automatic pipeline. Using these features, a proof-of-concept machine learning model is trained to predict sedimentation, achieving 84% accuracy with a 5cm error margin for a case study channel. The result is an automated node-based data processing pipeline using the FME software, alongside a geospatial database with 19 features for 26 side channels, and a machine learning model showcasing the predictive ability of the derived features.
- Open Access
- 11 April 2026
- Published by Atlantis Press SARL
- p. 102-117
Post-Colonial Wayfinding to the Landmark: Turns, Angles, and Visibility Across Time
- 11 April 2026
- Published by Atlantis Press SARL
- p. 102-117
Open AccessPost-Colonial Wayfinding to the Landmark: Turns, Angles, and Visibility Across Time
- Open Access
- 1 April 2026
- Published by KEAI Publishing in Frontiers of Architectural Research
Reframing the “H” in HBIM: from systematic review to UML-based conceptual modeling
- 1 April 2026
- Published by KEAI Publishing in Frontiers of Architectural Research
Open AccessReframing the “H” in HBIM: from systematic review to UML-based conceptual modeling
- Open Access
- 18 September 2025
- Published by Copernicus GmbH in The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
- Vol. XLVIII-4/W, 9-16
Point Clouds for 3D Land Administration: Integrating Floor Plans and Nationwide Airborne LiDAR (AHN)
As urban environments become increasingly vertical, Land Administration Systems (LAS) must support complex 3D spatial representations. While Building Information Models (BIM) offer such capabilities, they are not always available. This paper investigates an alternative approach using point clouds for 3D LAS, focusing on the integration of scanned cadastral floor plans and airborne LiDAR from the Actueel Hoogtebestand Nederland (AHN). We present a semi-automated pipeline that extracts floorplan geometries, segments and enhances AHN data, and synthesizes room-level point clouds. Results from a case study in Rotterdam demonstrate the potential of this approach in the absence of BIM, supporting legal space definition and public visualization. However, challenges such as misalignment due to occlusion in AHN data and inconsistent quality in older floor plan drawings affect the accuracy and automation of the process. The synthetic point clouds include room-level attributes, enabling a seamless integration with AHN, offering a representation of real-world features such as building facades, walls, and fences, which often delineate cadastral boundaries.
- 18 September 2025
- Published by Copernicus GmbH in The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
- Vol. XLVIII-4/W, 9-16
Open AccessPoint Clouds for 3D Land Administration: Integrating Floor Plans and Nationwide Airborne LiDAR (AHN)
As urban environments become increasingly vertical, Land Administration Systems (LAS) must support complex 3D spatial representations. While Building Information Models (BIM) offer such capabilities, they are not always available. This paper investigates an alternative approach using point clouds for 3D LAS, focusing on the integration of scanned cadastral floor plans and airborne LiDAR from the Actueel Hoogtebestand Nederland (AHN). We present a semi-automated pipeline that extracts floorplan geometries, segments and enhances AHN data, and synthesizes room-level point clouds. Results from a case study in Rotterdam demonstrate the potential of this approach in the absence of BIM, supporting legal space definition and public visualization. However, challenges such as misalignment due to occlusion in AHN data and inconsistent quality in older floor plan drawings affect the accuracy and automation of the process. The synthetic point clouds include room-level attributes, enabling a seamless integration with AHN, offering a representation of real-world features such as building facades, walls, and fences, which often delineate cadastral boundaries.
- Open Access
- 3 March 2024
- Published by Taylor & Francis in Journal of Location Based Services
- Vol. 19 (2) , 103-122
Indoor localisation through Isovist fingerprinting from point clouds and floor plans
- 3 March 2024
- Published by Taylor & Francis in Journal of Location Based Services
- Vol. 19 (2) , 103-122
Open AccessIndoor localisation through Isovist fingerprinting from point clouds and floor plans
- 21 February 2024
- Published by Springer Nature
- p. 329-334
Comparison of Cloud-to-Cloud Distance Calculation Methods - Is the Most Complex Always the Most Suitable?
- 21 February 2024
- Published by Springer Nature
- p. 329-334
Comparison of Cloud-to-Cloud Distance Calculation Methods - Is the Most Complex Always the Most Suitable?
- Open Access
- 6 June 2023
- Published by Copernicus GmbH in AGILE: GIScience Series
- Vol. 4, 1-9
Visualizing of the below-ground water network infrastructure
This research addresses the design and the utilization of vertical reference features in Augmented Reality (AR) to aid in providing perspective from underground pipes to the ground surface. The study utilizes 3D GIS data and assigns Z coordinates to the Underground Utility Networks (UUNs) using the AHN4 digital elevation data of the Netherlands (and either use standard depth or surveyed depth). However, registering Z coordinates and related data at the vertex level presents data registration issues that Esri's current Utility Network Model (UNM) cannot accommodate. A supplementary data model is required to address this, which can be achieved through vertical reference features. These multifunctional features emphasize the height difference between the pipe and the ground surface and can be used in AR applications to provide better depth clues. Additionally, polygons signatures and flow direction are utilized to enhance visualization.
- 6 June 2023
- Published by Copernicus GmbH in AGILE: GIScience Series
- Vol. 4, 1-9
Open AccessVisualizing of the below-ground water network infrastructure
This research addresses the design and the utilization of vertical reference features in Augmented Reality (AR) to aid in providing perspective from underground pipes to the ground surface. The study utilizes 3D GIS data and assigns Z coordinates to the Underground Utility Networks (UUNs) using the AHN4 digital elevation data of the Netherlands (and either use standard depth or surveyed depth). However, registering Z coordinates and related data at the vertex level presents data registration issues that Esri's current Utility Network Model (UNM) cannot accommodate. A supplementary data model is required to address this, which can be achieved through vertical reference features. These multifunctional features emphasize the height difference between the pipe and the ground surface and can be used in AR applications to provide better depth clues. Additionally, polygons signatures and flow direction are utilized to enhance visualization.
- Open Access
- 28 June 2021
- Published by Copernicus GmbH in The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
- Vol. XLIII-B3-2, 223-228
A COMPARATIVE STUDY OF POINT CLOUDS SEMANTIC SEGMENTATION USING THREE DIFFERENT NEURAL NETWORKS ON THE RAILWAY STATION DATASET
Point cloud data have rich semantic representations and can benefit various applications towards a digital twin. However, they are unordered and anisotropically distributed, thus being unsuitable for a typical Convolutional Neural Networks (CNN) to handle. With the advance of deep learning, several neural networks claim to have solved the point cloud semantic segmentation problem. This paper evaluates three different neural networks for semantic segmentation of point clouds, namely PointNet++, PointCNN and DGCNN. A public indoor scene of the Amersfoort railway station is used as the study area. Unlike the typical indoor scenes and even more from the ubiquitous outdoor ones in currently available datasets, the station consists of objects such as the entrance gates, ticket machines, couches, and garbage cans. For the experiment, we use subsets from the data, remove the noise, evaluate the performance of the selected neural networks. The results indicate an overall accuracy of more than 90% for all the networks but vary in terms of mean class accuracy and mean Intersection over Union (IoU). The misclassification mainly occurs in the classes of couch and garbage can. Several factors that may contribute to the errors are analyzed, such as the quality of the data and the proportion of the number of points per class. The adaptability of the networks is also heavily dependent on the training location: the overall characteristics of the train station make a trained network for one location less suitable for another.
- 28 June 2021
- Published by Copernicus GmbH in The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
- Vol. XLIII-B3-2, 223-228
Open AccessA COMPARATIVE STUDY OF POINT CLOUDS SEMANTIC SEGMENTATION USING THREE DIFFERENT NEURAL NETWORKS ON THE RAILWAY STATION DATASET
Point cloud data have rich semantic representations and can benefit various applications towards a digital twin. However, they are unordered and anisotropically distributed, thus being unsuitable for a typical Convolutional Neural Networks (CNN) to handle. With the advance of deep learning, several neural networks claim to have solved the point cloud semantic segmentation problem. This paper evaluates three different neural networks for semantic segmentation of point clouds, namely PointNet++, PointCNN and DGCNN. A public indoor scene of the Amersfoort railway station is used as the study area. Unlike the typical indoor scenes and even more from the ubiquitous outdoor ones in currently available datasets, the station consists of objects such as the entrance gates, ticket machines, couches, and garbage cans. For the experiment, we use subsets from the data, remove the noise, evaluate the performance of the selected neural networks. The results indicate an overall accuracy of more than 90% for all the networks but vary in terms of mean class accuracy and mean Intersection over Union (IoU). The misclassification mainly occurs in the classes of couch and garbage can. Several factors that may contribute to the errors are analyzed, such as the quality of the data and the proportion of the number of points per class. The adaptability of the networks is also heavily dependent on the training location: the overall characteristics of the train station make a trained network for one location less suitable for another.
- Open Access
- 17 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 44150-44161
An Approach to Map Visibility in the Built Environment From Airborne LiDAR Point Clouds
Sustainable development can only be achieved with an innovative improvement from the way we currently analyze, design, build and manage our urban spaces. Current digital analysis and design methods for cities, such as visibility analysis, deeply rely on mapping and modeling techniques. However, most methods fall short of depicting the real visual landscape in the urban realm and this could bring a significant error in visibility calculations which may lead to an improper decision for urban spaces. The technical development of light detection and ranging(LiDAR) technology introduces new approaches for urban study. LiDAR utilizes point clouds including thousands or even millions of georeferenced points, and thus can support 3-D digital representation of urban landscape with detailed information and high resolution. Besides the superiority in representing urban landscape, LiDAR point clouds also has a clear advantage in quantitative analysis and provides better visibility than traditional models. In this paper, we first introduced a novel approach to map visibility in the urban built environment involving vegetation data directly using airborne LiDAR point clouds. This approach calculates neighborhood statistics for occlusion detection. Then we presented 2 case with different scenarios showing how our approach can be used to obtain a precise visibility in an urban area in the Netherlands. At last, we discussed how point clouds based visibility models can be further explored and can better assist urban design.
- 17 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 44150-44161
Open AccessAn Approach to Map Visibility in the Built Environment From Airborne LiDAR Point Clouds
Sustainable development can only be achieved with an innovative improvement from the way we currently analyze, design, build and manage our urban spaces. Current digital analysis and design methods for cities, such as visibility analysis, deeply rely on mapping and modeling techniques. However, most methods fall short of depicting the real visual landscape in the urban realm and this could bring a significant error in visibility calculations which may lead to an improper decision for urban spaces. The technical development of light detection and ranging(LiDAR) technology introduces new approaches for urban study. LiDAR utilizes point clouds including thousands or even millions of georeferenced points, and thus can support 3-D digital representation of urban landscape with detailed information and high resolution. Besides the superiority in representing urban landscape, LiDAR point clouds also has a clear advantage in quantitative analysis and provides better visibility than traditional models. In this paper, we first introduced a novel approach to map visibility in the urban built environment involving vegetation data directly using airborne LiDAR point clouds. This approach calculates neighborhood statistics for occlusion detection. Then we presented 2 case with different scenarios showing how our approach can be used to obtain a precise visibility in an urban area in the Netherlands. At last, we discussed how point clouds based visibility models can be further explored and can better assist urban design.