Floods are one of the most serious natural disasters that can cause huge societal and economic losses.Extensive research has been conducted on topics like flood monitoring,prediction,and loss estimation.In these resea...Floods are one of the most serious natural disasters that can cause huge societal and economic losses.Extensive research has been conducted on topics like flood monitoring,prediction,and loss estimation.In these research fields,flood velocity plays a crucial role and is an important factor that influences the reliability of the outcomes.Traditional methods rely on physical models for flood simulation and prediction and could generate accurate results but often take a long time.Deep learning technology has recently shown significant potential in the same field,especially in terms of efficiency,helping to overcome the time-consuming associated with traditional methods.This study explores the potential of deep learning models in predicting flood velocity.More specifically,we use a Multi-Layer Perceptron(MLP)model,a specific type of Artificial Neural Networks(ANNs),to predict the velocity in the test area of the Lundesokna River in Norway with diverse terrain conditions.Geographic data and flood velocity simulated based on the physical hydraulic model are used in the study for the pre-training,optimization,and testing of the MLP model.Our experiment indicates that the MLP model has the potential to predict flood velocity in diverse terrain conditions of the river with acceptable accuracy against simulated velocity results but with a significant decrease in training time and testing time.Meanwhile,we discuss the limitations for the improvement in future work.展开更多
Reliable and up-to-date geospatial data plays a fundamental role in Sustainable Development Goals(SDGs)monitoring.Aiming to providing such geospa-tial data,numerous algorithms,solutions and frame-works have been devel...Reliable and up-to-date geospatial data plays a fundamental role in Sustainable Development Goals(SDGs)monitoring.Aiming to providing such geospa-tial data,numerous algorithms,solutions and frame-works have been developed in recent years.A mong oth-ers,A rtificial Intelligence(AI)based techniques have been widely used for the tasks of processing geospatial data.Nowadays,this topic is blooming so fast and to a vast extent in the field of Geomatics that a new subdo-main seems to arise,namely GeoAI[1-2].Even for a very quick and brief glance inthe Internet,people can find a lot of applications,projects,blogs and research articles about GeoAI,w hereas new approaches to GeoAI have been proposed and tested.展开更多
The OGC standard for 3D city modeling is widely used in an increasing number of applications. It defines five consecutive Levels of Detail （LoD0 to LoD4 with increasing accuracy and structural complexity）, in which ...The OGC standard for 3D city modeling is widely used in an increasing number of applications. It defines five consecutive Levels of Detail （LoD0 to LoD4 with increasing accuracy and structural complexity）, in which LoD3 includes all exterior appearances and geometrical details and subsequently requires much storage space. A new LoD is introduced as shell model with the exterior shell of the LoD3 model and the opening objects like windows, doors as well as smaller facade objects are projected onto walls. In this paper, a user survey is presented. The results of this survey show that the shell model can give users almost the same visual impression as the LoD3 model. Furthermroe, algorithms are developed to extract the shell model from LoD3 model. Experiments show that this shell model can reduce up to 90% storage of the original LoD3 model. Therefore, on one hand it can be used as a substitute for a LoD3 model for the visualization on small displays. On the other hand, it can be treated as a sub-level of detail （SLoD3） in CityGML, since it retains almost the same amount of information but requires much less storage space.展开更多
文摘Floods are one of the most serious natural disasters that can cause huge societal and economic losses.Extensive research has been conducted on topics like flood monitoring,prediction,and loss estimation.In these research fields,flood velocity plays a crucial role and is an important factor that influences the reliability of the outcomes.Traditional methods rely on physical models for flood simulation and prediction and could generate accurate results but often take a long time.Deep learning technology has recently shown significant potential in the same field,especially in terms of efficiency,helping to overcome the time-consuming associated with traditional methods.This study explores the potential of deep learning models in predicting flood velocity.More specifically,we use a Multi-Layer Perceptron(MLP)model,a specific type of Artificial Neural Networks(ANNs),to predict the velocity in the test area of the Lundesokna River in Norway with diverse terrain conditions.Geographic data and flood velocity simulated based on the physical hydraulic model are used in the study for the pre-training,optimization,and testing of the MLP model.Our experiment indicates that the MLP model has the potential to predict flood velocity in diverse terrain conditions of the river with acceptable accuracy against simulated velocity results but with a significant decrease in training time and testing time.Meanwhile,we discuss the limitations for the improvement in future work.
文摘Reliable and up-to-date geospatial data plays a fundamental role in Sustainable Development Goals(SDGs)monitoring.Aiming to providing such geospa-tial data,numerous algorithms,solutions and frame-works have been developed in recent years.A mong oth-ers,A rtificial Intelligence(AI)based techniques have been widely used for the tasks of processing geospatial data.Nowadays,this topic is blooming so fast and to a vast extent in the field of Geomatics that a new subdo-main seems to arise,namely GeoAI[1-2].Even for a very quick and brief glance inthe Internet,people can find a lot of applications,projects,blogs and research articles about GeoAI,w hereas new approaches to GeoAI have been proposed and tested.
基金the National Natural Science Foundation of China(No. 41071288)
文摘The OGC standard for 3D city modeling is widely used in an increasing number of applications. It defines five consecutive Levels of Detail （LoD0 to LoD4 with increasing accuracy and structural complexity）, in which LoD3 includes all exterior appearances and geometrical details and subsequently requires much storage space. A new LoD is introduced as shell model with the exterior shell of the LoD3 model and the opening objects like windows, doors as well as smaller facade objects are projected onto walls. In this paper, a user survey is presented. The results of this survey show that the shell model can give users almost the same visual impression as the LoD3 model. Furthermroe, algorithms are developed to extract the shell model from LoD3 model. Experiments show that this shell model can reduce up to 90% storage of the original LoD3 model. Therefore, on one hand it can be used as a substitute for a LoD3 model for the visualization on small displays. On the other hand, it can be treated as a sub-level of detail （SLoD3） in CityGML, since it retains almost the same amount of information but requires much less storage space.
