Ground and building extraction from LiDAR data based on differential morphological profiles and locally fitted surfaces

Mongus, Domen; Lukač, Niko; Žalik, Borut

doi:10.1016/j.isprsjprs.2013.12.002

Cited by 165 publications

(86 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…During the last decades, many filtering algorithms have been explored and developed for classifying top-view LiDAR point cloud in order to extract some key components of urban features, e.g. land covers (Yan et al, 2015), trees (Alonzo et al, 2014;Han et al, 2014;Chen et al, 2015), buildings (Kabolizade et al, 2010;Awrangjeb et al, 2013;Mongus et al, 2014;Song et al, 2015;Ferraz et al, 2016), roads (Li et al, 2015;Ferraz et al, 2016), or even vehicles (Yao et al, 2010). When a set of criteria has been characterised, essential information embedded in point cloud can be extracted and classified into particular segments.…”

Section: Top-view Lidar Point Cloud Extractionmentioning

confidence: 99%

Point Cloud Data Fusion for Enhancing 2D Urban Flood Modelling

Meesuk¹

2017

View full text Add to dashboard Cite

Although all care is taken to ensure the integrity and the quality of this publication and the information herein, no responsibility is assumed by the publishers, the author nor UNESCO-IHE for any damage to the property or persons as a result of operation or use of this publication and/or the information contained herein. A pdf version of this work will be made available as Open Access via http://repository.tudelft.nl/ihe This version is licensed under the Creative Commons Attribution-Non Commercial 4.0 International License, http://creativecommons.org/licenses/by-nc/4.0/Published by: CRC Press/Balkema PO Box 11320, 2301 EH Leiden, the Netherlands e-mail: Pub.NL@taylorandfrancis.com www.crcpress.com -www.taylorandfrancis.com to my dear family SummaryModelling urban flood dynamics requires detailed knowledge of a number of complex processes. Numerical simulation of flood propagation requires proper understanding of all processes involved. Also, special care has to be taken to adequately represent the urban topography, taking into account typical urban features like underpasses and hidden alleyways. Incorrect input data will affect the numerical model results, which may lead to inadequate flood-protection measures or even catastrophic flooding situations. This thesis explores how to include particular urban topographic features in urban flood modelling.Aerial Light Detection and Ranging (LiDAR) systems offer opportunities for achieving good quality topographic data, with less fieldwork on the ground. Even though aerial LiDAR data have long been used in many applications, conventional top-view LiDAR data can not quite capture some hidden urban features. However, recent improvements in Structure from Motion (SfM) techniques provide opportunities to achieve improved quality topographic data by using multiple viewpoints, including side-view data.This dissertation explores insights into the capabilities of assimilating SfM point cloud data by using multi-source views as input for enhancing 2D urban flood modelling. Side-view SfM point cloud data are collected and merged with conventional top-view LiDAR point cloud data to create novel Multi-Source Views (MSV) topographic data. The main objectives of this research are (i) to provide insight into the capabilities of using computer-based environments; (ii) to explore the benefits of using the new MSV data; (iii) to enhance 2D model schematizations; (iv) to compare simulated results using new MSV data and conventional top-view LiDAR data as input for urban flood models; and (v) to help developing flood-protection measures. Findings showed that simulation results using conventional top-view LiDAR-DSM as input show the least flood inundation areas and results contained many dry areas. This is because conventional LiDAR-DSM does not capture hidden urban features like underpasses, high trees, overarching structures, which behave as obstacles in 2D model schematics. When applying extended top-view LiDAR-DBM+ and the newly developed MSV-DEM as input, simulation results showed m...

show abstract

Section: Top-view Lidar Point Cloud Extractionmentioning

confidence: 99%

Point Cloud Data Fusion for Enhancing 2D Urban Flood Modelling

Meesuk¹

2017

View full text Add to dashboard Cite

show abstract

“…It is defined as the distance between the studied point Pi and the least square best fitting plane comprising Pi and its neighborhood points inside the radius r sphere [32]. This is similar to the Locally Fitted Surfaces (LoFS) implemented in [33]. The lowest roughness values correspond to flat surfaces while higher roughness values take place in those elements with irregular shapes [16].…”

Section: Roughnessmentioning

confidence: 99%

Morphological Operations to Extract Urban Curbs in 3D MLS Point Clouds

Rodríguez-Cuenca

García-Cortés

Ordóñez

et al. 2016

IJGI

View full text Add to dashboard Cite

Automatic curb detection is an important issue in road maintenance, three-dimensional (3D) urban modeling, and autonomous navigation fields. This paper is focused on the segmentation of curbs and street boundaries using a 3D point cloud captured by a mobile laser scanner (MLS) system. Our method provides a solution based on the projection of the measured point cloud on the XY plane. Over that plane, a segmentation algorithm is carried out based on morphological operations to determine the location of street boundaries. In addition, a solution to extract curb edges based on the roughness of the point cloud is proposed. The proposed method is valid in both straight and curved road sections and applicable both to laser scanner and stereo vision 3D data due to the independence of its scanning geometry. The proposed method has been successfully tested with two datasets measured by different sensors. The first dataset corresponds to a point cloud measured by a TOPCON sensor in the Spanish town of Cudillero. The second dataset corresponds to a point cloud measured by a RIEGL sensor in the Austrian town of Horn. The extraction method provides completeness and correctness rates above 90% and quality values higher than 85% in both studied datasets.

show abstract

“…In the third group of modelling approaches, the combination of model driven and data driven algorithms is used to have an optimal method for compensating the weakness of each methods. According to our study, current methodologies and algorithms on building detection and extraction problem can be divided into four groups as; plan fitting based methods (Mongus, et al, 2014); filtering and thresholding based methods (Maltezos, et al, 2015;Hermosilla, et al, 2011) such as morphological methods (Yu, et al, 2010); segmentation based methods such as binary space partitioning (Wichmann, et al, 2015), shadow based segmentation (Singh, et al, 2015;Ngo, et al, 2015), and region growing based algorithms (Matikainen, et al, 2010;Awrangjeb, et al, 2013); and finally the latest group, different supervised classification methods (Hermosilla, et al, 2011;Guo, et al, 2011;Karantzalos, et al, 2015;Vakalopoulou, et. al, 2011).…”

Section: Introductionmentioning

confidence: 99%

Knowledge Based 3d Building Model Recognition Using Convolutional Neural Networks From Lidar and Aerial Imageries

Alidoost¹,

Arefi²

2016

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:In recent years, with the development of the high resolution data acquisition technologies, many different approaches and algorithms have been presented to extract the accurate and timely updated 3D models of buildings as a key element of city structures for numerous applications in urban mapping. In this paper, a novel and model-based approach is proposed for automatic recognition of buildings' roof models such as flat, gable, hip, and pyramid hip roof models based on deep structures for hierarchical learning of features that are extracted from both LiDAR and aerial ortho-photos. The main steps of this approach include building segmentation, feature extraction and learning, and finally building roof labeling in a supervised pre-trained Convolutional Neural Network (CNN) framework to have an automatic recognition system for various types of buildings over an urban area. In this framework, the height information provides invariant geometric features for convolutional neural network to localize the boundary of each individual roofs. CNN is a kind of feed-forward neural network with the multilayer perceptron concept which consists of a number of convolutional and subsampling layers in an adaptable structure and it is widely used in pattern recognition and object detection application. Since the training dataset is a small library of labeled models for different shapes of roofs, the computation time of learning can be decreased significantly using the pre-trained models. The experimental results highlight the effectiveness of the deep learning approach to detect and extract the pattern of buildings' roofs automatically considering the complementary nature of height and RGB information.

show abstract

Ground and building extraction from LiDAR data based on differential morphological profiles and locally fitted surfaces

Cited by 165 publications

References 37 publications

Point Cloud Data Fusion for Enhancing 2D Urban Flood Modelling

Point Cloud Data Fusion for Enhancing 2D Urban Flood Modelling

Morphological Operations to Extract Urban Curbs in 3D MLS Point Clouds

Knowledge Based 3d Building Model Recognition Using Convolutional Neural Networks From Lidar and Aerial Imageries

Contact Info

Product

Resources

About