The energy consumption and emissions from vehicles adversely affect human health urban sustainability. Analysis of GPS big data collected can provide useful insights about the quantity distribution such emissions. Previous studies, which estimated fuel consumption/emissions traffic based on sampled data, have not sufficiently considered vehicle activities may led to erroneous estimations. By adopting analytical construct space-time path in time geography, this study proposes methods that...

Detailed real-time road data are an important prerequisite for navigation and intelligent transportation systems. As accident-prone areas, intersections play a critical role in route guidance traffic management. Ubiquitous trajectory have led to recent surge map reconstruction. However, it is still challenging automatically generate detailed structural models intersections, especially from low-frequency data. We propose novel three-step approach extract the semantic information of...

Short-term traffic prediction is of great importance to the management congestion, a pervasive and difficult-to-solve problem in many metropolises all over world. However, existing studies on contain rough information at carriageway level that ignore distinction between different turns one intersection. With aim predicting road intersections from big trace data finer scale, this study proposes novel method, fine-grained method (FTPG) with graph attention network (GAT), which predicts...

Detailed and precise urban green spaces (UGS) maps provide essential data for the sustainable development related studies (e.g. heatwave events, heat health risk, flooding, biodiversity ecosystem services). However, remote sensing of mapping UGS is challenging due to existence mixed pixels cost difficulty collecting quality training data. This study presents a neural network-based automatic method that integrates use Sentinel-2A satellite images crowdsourced geospatial big The proposed...

In this paper, a bottom–up vehicle emission model is proposed to estimate real-time <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$\hbox{CO}_{2}$</tex></formula> emissions using intelligent transportation system (ITS) technologies. the model, traffic data that were collected by ITS are fully utilized detailed technology (e.g., type) and driving pattern speed, acceleration, road slope) in network....

Lane-based road network information, such as the number and locations of traffic lanes on a road, has played an important role in intelligent transportation systems. In this paper, we propose Collecting Road Information via Crowdsourcing (CLRIC) method, which can automatically extract detailed lane structure roads by using crowdsourcing data collected vehicles. First, CLRIC filters high-precision GPS from raw trajectories based region growing clustering with prior knowledge. Second, mines...

In this paper, we propose a novel approach for mining lane-level road network information from low-precision vehicle GPS trajectories (MLIT), which includes the number and turn rules of traffic lanes based on naïve Bayesian classification. First, proposed method (MLIT) uses an adaptive density optimization to remove outliers raw their space-time distribution clustering. Second, MLIT acquires in two steps. The first step establishes classifier according trace features plane profiles real...

The presence of clouds greatly reduces the ground information high-resolution satellite data. In order to improve utilization data, this article presents a cloud removal method based on deep learning. This is first end-to-end architecture that has great potential detect and remove from For detection, convolution neural network (CNN) used them. removal, content generation network, texture spectrum traditional CNN are proposed. proposed can use multisource data (content, texture, spectral) as...

Lane-level road network updating is crucial for urban traffic applications that use geographic information systems contributing to, example, intelligent driving, route planning and control. Researchers have developed various algorithms to update networks using sensor data, such as high-definition images or GPS data; however, approaches involve change detection at lane level data are less common. This paper presents a novel method automatic of lane-level based on trajectories vehicles. The...

Kernel Density Estimation (KDE) is an important approach to analyse spatial distribution of point features and linear over 2-D planar space. Some network-based KDE methods have been developed in recent years, which focus on estimating density events 1-D network However, the existing are not appropriate for analysing characteristics certain kind or events, such as traffic jams, queue at intersections taxi carrying passenger events. These occur distribute road space, present a continuous along...

With the rapid development of urban transportation, people urgently need high-precision and up-to-date road maps. At same time, themselves are an important source information for detailed map construction, as they can detect real-world surfaces with GPS devices in course their everyday life. Big trace data makes it possible provides a great opportunity to extract refine maps at relatively low cost. In this paper, new refinement method is proposed incremental construction using big data,...

Landsat images have played an important role in the field of Earth observation and geoinformatics. However, optical are frequently contaminated by cloud cover, especially tropical subtropical regions, which limits utilization these images. To improve images, this study, we propose a novel spatiotemporal neural network with four modules: detection module, spatial–temporal learning feature fusion reconstruction module. The results experiments demonstrate that proposed method is quantitatively...

Cloud and snow detection is one of the most significant tasks for remote sensing image processing. However, it a challenging task to distinguish between clouds in high-resolution multispectral images due their similar spectral distributions. The shortwave infrared band (SWIR, e.g., Sentinel-2A 1.55–1.75 µm band) widely applied clouds. have lack SWIR, such traditional methods are no longer practical. To solve this problem, novel convolutional neural network (CNN) classify cloud on an object...

ABSTRACTIntersections are the critical parts where different traffic flows converge and change directions, forming "bottlenecks" "clog points" in urban traffic. Intersection travel time is an important parameter for public route planning, management, engineering optimization. Based on low-frequency spatial-temporal Global Positioning System (GPS) trace data, this article presents a novel method estimating intersection time. The proposed first analyzes patterns of vehicles through...

Cloud detection is a crucial procedure in remote sensing preprocessing. However, cloud challenging cloud–snow coexisting areas because and snow have similar spectral characteristic visible spectrum. To overcome this challenge, we presented an automatic neural network (ACD net) integrated imagery with geospatial data aimed to improve the accuracy of from high-resolution under coexistence. The proposed ACD net consisted two parts: 1) feature extraction networks 2) boundary refinement module....

Cloud detection is a fundamental step for optical satellite image applications. Although existing deep learning method can provide more accurate cloud result. However, performance of these methods rely on large number label samples, whose collection time-consuming and high cost. In addition, challenging in brightness scenes due to object have similar spectral feature. this study, we propose index driven spectral-spatial-context attention network (SSCA-net) detection, which relies no effort...