Road pavement damage affects driving comfort markedly, threatens safety, and may even cause traffic accidents. The management department conventionally captures information mainly using manual vehicle-mounted equipment, which is not conducive to the detection of large-scale road distress. Street-view images can provide full-view urban roads where data updated regularly by navigation map service companies, making it possible rapidly detect in areas. This paper presents a new approach that...

The UN Sustainable Development Goals allude to the importance of infrastructure quality in three its seventeen goals. However, monitoring developing regions remains prohibitively expensive and impedes efforts measure progress toward these To this end, we investigate use widely available remote sensing data for prediction Africa. We train a convolutional neural network predict ground truth labels from Afrobarometer Round 6 survey using Landsat 8 Sentinel 1 satellite imagery. Our best models...

Hydrogen is an attractive, clean, sustainable energy source primarily produced via industry. At present, most reviews on hydrogen mainly focus the preparation and storage of hydrogen, while development utilization natural will greatly reduce its cost. Natural has been discovered in many geological environments. Therefore, based extensive literature research, this study, distribution sources were systematically sorted, identification method occurrence state examined summarized. The results...

Conventional methods for monitoring pavement healthy states have the disadvantages of low efficiency and being time-consuming destructive. Current studies indicate that traditional machine learning algorithms showed poor performance generalization capacity in identifying asphalt aging damage conditions. Further, deep network models less been applied to detection types objects from UAV imagery. In this study, we first used a low-altitude platform acquire multispectral images road with...

Road pavement cracks are a critical factor affecting the health conditions of road pavements. Accurate crack detection contributes to providing data support for maintenance measures. Compared conventional algorithms, deep learning based segmentation methods have practical significance and traffic safety management due their high precision automation. However, existing often suffer from accuracy loss varying sizes presence stains on with spatial characteristics similar cracks, leading...

Transformer-based deep networks have been widely employed for fusing multispectral and hyperspectral images because the global self-attention mechanism ensures that more pixel information is used in fusion. However, unreasonable interactions between local windows window self-attention-based Transformer can adversely affect network's modeling of relationship, resulting low-quality fusion results. This study proposes a novel N-Gram based Swin Network (NGST-Net). It utilizes pixels mining...

The automated monitoring of road pavement conditions is a challenging subject in intelligent transportation. However, the existing studies mostly focus on extracting damages such as cracks, while aging are still less investigated. In this paper, novel method based modified recurrent neural network designed for asphalt phenomena from fine-resolution satellite imagery. A spectral augmentation proposed to enhance details pavements. loss function also improve bi-directional gated unit (Bi-GRU)...

Deep learning has been developed to generate promising super resolution hyperspectral imagery by fusing with the panchromatic band. However, it is still challenging maintain edge spectral information in necessary upsampling processes of these approaches, and difficult guarantee effective feature extraction. This study proposes a pansharpening network denoted as HyperRefiner that consists of, (1) well performing SRNet, which dual attention block refined are cascaded accomplish extraction...

The autoencoder (AE) framework is usually adopted as a baseline network for hyperspectral unmixing. Totally AE performs well in unmixing through automatically learning low-dimensional embedding and reconstructing data. However, most available AE-based networks do not fully consider the spatial spectral information of different ground features images output relatively fixed ratios linear nonlinear photon scattering effects under scenarios. Therefore, these methods have poor generalization...

Early detection of people’s suspicious behaviors can aid in the prevention crimes and make community safer. Existing methods that are focused on identifying abnormal from video surveillance based computer vision, which more suitable for detecting ongoing behaviors. While criminals intend to avoid under surveillance, their prior will be unconsciously reflected trajectories. Herein, we characterize several unusual movement patterns, behaviors, gatherings people, analyze features hidden...

Many real-world complex systems consist of a set basic units that are connected by different kinds relationships. All types such can be described multilayer network, where each link represents interaction among the same nodes. In this paper, we present general framework to characterize influences (centrality) layers. Furthermore, propose two measures for layer centrality in terms network connectivity under framework. The idea our consists assigning more value layers contribute network. other...

How to predict the future popularity of a message or video on online social media (OSM) has long been an attractive problem for researchers. Although many difficulties are still ahead, recent studies suggest that temporal and topological features early adopters generally play very important role. However, with increase adopters, feature space will grow explosively. select most effective is open issue. In this work, we investigate several extraction methods over Twitter platform find...

Detailed snow cover maps are essential for estimating the earth’s energy balance and hydrological cycle. Mapping across spatially extensive topographically complex areas with less or no cloud obscuration is challenging, but SNPP-VIIRS Day/Night Band (DNB) nighttime light data offers a potential solution. This paper aims to map distribution at 750 m resolution diverse 1,664,900 km2 of Xinjiang, China, based on DNB radiance. We implemented swarm intelligent optimization technique Krill Herd...

This paper presents two improved techniques to determine boundaries and depth from observed gravity or magnetic anomalies. The first technique is based on analysis of the largest curvature total horizontal gradient field boundaries. second signal estimate depth. just only calculate magnitude anomalies, rather than derivatives magnitude. It a particularly useful transformation for reducing effects noise increasing coherency solutions model-independent functions. shown work successfully in...

In order to explore the performance of B-ULV-616A knapsack sprayer, computational fluid dynamics (CFD) was used simulate air-assisted device, which features an ultra-low-volume electric sprayer. Field experiments were carried out test spraying effects, and KANOMAX anemometer verify simulated results. First, internal external flow fields droplet deposition distribution sprayer established. The results showed that spray device can change airflow speed direction produce a high-speed swirling at...

This paper mainly discusses how to determine the optimized strategy for retrieval of Precipitable Water Vapor (PWV) with high accuracy from tropospheric zenith wet delay using ground-based GPS receivers. analytical network are constructed on base two observation sites in Antarctica 1999 and several IGS sites. Tests conducted study performance different sizes schemes parameters. A high-accuracy processing software package GAMIT/GLOBK is utilized; multiple adopted searching parameters accurate...

Atomizers were designed with different atomization parameters to obtain droplets that satisfy optimal particle size requirements for an impinging-type low-speed centrifugal atomizing sprayer. The main factors affecting droplet are turntable speed, the number of teeth and tooth shape toothed disc. Winner318 software was used evaluate sizes structures working atomizer. response surface method Design-Expert analyze effect each factor. analysis structural atomizer on interaction between volume...