Procedural generation of flood-sensitive urban layouts

Mustafà, Ahmed; Zhang, Xiao Wei; Aliaga, Daniel G.; Bruwier, Martin; Nishida, Gen; Dewals, Benjamin; Erpicum, Sébastien; Archambeau, Pierre; Pirotton, Michel; Teller, Jacques

doi:10.1177/2399808318812458

Cited by 22 publications

(26 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Over the past two decades, MMSs have been developed and used in several fields, such as urban planning [12], 3D building modeling [13][14][15][16], virtual heritage conservation [17], augmented reality [18], transportation [19,20], and forestry [21]. Recent applications of 3D urban modeling are actively used in the field of disaster management and flood-sensitive cities [22].…”

Section: Related Workmentioning

confidence: 99%

“…Generally, MMSs are used to fulfill demands for high-quality data collection in a manner that is usually not possible when using airborne surveys to supply comprehensible city models. While the terrains, outlines, and roof shapes of buildings can be reconstructed using aerial surveys, there is only limited information on building facades available for data processing from this source [8,18,22,36].…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Developing a Mobile Mapping System for 3D GIS and Smart City Planning

Yang

2019

Sustainability

View full text Add to dashboard Cite

The creation of augmented reality-related geographic information system (GIS) mapping applications has witnessed considerable advances in the technology of urban modeling; however, there are limitations to the technology that is currently used to create similar resources. The cost of the creation of the vehicle is an obstacle, and the rendering of textures of buildings is often lacking because of the distortion caused by the types of lenses that have been used. Generally, mobile mapping systems (MMSs) can extract detailed three-dimensional (3D) data with high quality texture information of the 3D building model. However, mapping urban areas by MMSs is expensive and requires advanced mathematical approaches with complicated steps. In particular, commercial MMS, which generally includes two GPS receivers, is an expensive device, costing ~$1 million. Thus, this research is aimed at developing a new MMS that semi-automatically produces high-quality texture information of 3D building models proposes a 3D urban model by hybrid approaches. Eventually, this study can support urban planners and people to improve their spatial perception and awareness of urban area for Smart City Planning.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Developing a Mobile Mapping System for 3D GIS and Smart City Planning

Yang

2019

Sustainability

View full text Add to dashboard Cite

show abstract

“…For example, regression forest was used to learn the dynamic process of particle‐based simulations for real‐time animation and layout planning (Feng, Yu, Yeung, Yin, & Zhou, 2016; Ladický, Jeong, Solenthaler, Pollefeys, & Gross, 2015). Later, a fully‐connected neural network was proposed by Mustafa et al (2018) for flood‐driven urban planning. However, the approach had a main drawback that the training data for the neural network was the input parameters of a terrain generator rather than raw elevation data, and consequently the results cannot be extended to other scenarios.…”

Section: Introductionmentioning

confidence: 99%

Data‐driven flood emulation: Speeding up urban flood predictions by deep convolutional neural networks

Guo

Leitão

Simões

et al. 2020

J Flood Risk Management

View full text Add to dashboard Cite

Computational complexity has been the bottleneck for applying physically based simulations in large urban areas with high spatial resolution for efficient and systematic flooding analyses and risk assessment. To overcome the issue of long computational time and accelerate the prediction process, this paper proposes that the prediction of maximum water depth can be considered an image‐to‐image translation problem in which water depth rasters are generated using the information learned from data instead of by conducting simulations. The proposed data‐driven urban pluvial flood approach is based on a deep convolutional neural network trained using flood simulation data obtained from three catchments and 18 hyetographs. Multiple tests to assess the accuracy and validity of the proposed approach were conducted with both design and real hyetographs. The results show that flood prediction based on neural networks use only 0.5% of the time compared with that of physically based models, with promising accuracy and generalizability. The proposed neural network can also potentially be applied to different but relevant problems, including flood analysis for flood‐safe urban layout planning.

show abstract

“…The initial form of LiDAR data, as represented by an unorganized 3D point cloud, has many disadvantages, including poor support of representing objects and their attributes, low efficiency with respect to used processing power and memory resources (as even primitive surfaces, such as flat building roofs, are represented by dense groups of points), and low visualization quality. As a result, certain areas of research prefer the use of simpler procedurally generated models [22]. For these reasons, a different type of geometric model represents 3D spatial data, which relies on expressing three-dimensional objects in the form of surfaces consisting of higher order geometric structures, known as triangulated irregular network (TIN) models.…”

Section: Introductionmentioning

confidence: 99%

Integration, Processing and Dissemination of LiDAR Data in a 3D Web-GIS

Kulawiak

Łubniewski

2019

IJGI

View full text Add to dashboard Cite

The rapid increase in applications of Light Detection and Ranging (LiDAR) scanners, followed by the development of various methods that are dedicated for survey data processing, visualization, and dissemination constituted the need of new open standards for storage and online distribution of collected three-dimensional data. However, over a decade of research in the area has resulted in a number of incompatible solutions that offer their own ways of disseminating results of LiDAR surveys (be it point clouds or reconstructed three-dimensional (3D) models) over the web. The article presents a unified system for remote processing, storage, visualization, and dissemination of 3D LiDAR survey data, including 3D model reconstruction. It is built with the use of open source technologies and employs open standards, such as 3D Tiles, LASer (LAS), and Object (OBJ) for data distribution. The system has been deployed for automatic organization, processing, and dissemination of LiDAR surveys that were performed in the city of Gdansk. The performance of the system has been measured using a selection of LiDAR datasets of various sizes. The system has shown to considerably simplify the process of data organization and integration, while also delivering tools for easy discovery, inspection, and acquisition of desired datasets.

show abstract

Procedural generation of flood-sensitive urban layouts

Cited by 22 publications

References 56 publications

Developing a Mobile Mapping System for 3D GIS and Smart City Planning

Developing a Mobile Mapping System for 3D GIS and Smart City Planning

Data‐driven flood emulation: Speeding up urban flood predictions by deep convolutional neural networks

Integration, Processing and Dissemination of LiDAR Data in a 3D Web-GIS

Contact Info

Product

Resources

About