Smartphone-based collaborative indoor localization technology is a rapidly evolving area of research that aims to provide convenient and accurate user indoor location information. This technology encourages users to collaborate with each other using sensors commonly found in smartphones, such as WiFi, Bluetooth, and inertial sensors. By collecting and sharing data from multiple users, the collaborative system can overcome the limitations of individual devices and improve localization accuracy and efficiency. However, the need for fingerprint database establishment, frequent calibration, and the effects of signal strength fluctuations still plagues the development of the technology. To address these limitations, this paper proposes a collaborative indoor localization scheme that combines smartphone WiFi hotspots for information transmission with inertial sensor data to improve localization accuracy. The proposed method does not require additional equipment or tedious fingerprint database collection, and the use of information on the trend of WiFi signal strength changes also reduces the impact of signal fluctuations to some extent. The collaborative scheme can estimate the relative position and exchange information in an environment without external devices, which is of great significance for collaborative work and emergency response in special situations. Extensive experiments conducted in our office area demonstrate the proposed solution’s ability to achieve satisfactory localization results.
Total phosphorus mapping forms the foundation of water pollution prevention and control, constituting a pivotal means for water environment governance. The Google Earth Engine (GEE) cloud platform boasts an extensive collection of remote sensing imagery data and substantial computational prowess. This platform presents an opportunity to furnish robust support for total phosphorus monitoring, bypassing the need for extensive data downloads, local storage, and processing. In this paper, we employ Sentinel-2 remote sensing data to extract water bodies, specifically focusing on the Jiujiang area in Jiangxi Province, China. Drawing from both water body data and monitoring total phosphorus obtained through traditional ground stations, we construct a model for the inversion of total phosphorus (TP) concentrations. The proposed model attains an R 2 value of 0.42.This model's precision is also assessed. Furthermore, we employ the proposed model and GEE to derive TP concentrations for Jiujiang's water bodies in October 2022. Inference from the inversion outcomes highlights that Poyang Lake exhibits a TP concentration exceeding 0.1 mg/L, surpassing that of other regions. Consequently, the total phosphorus model and inversion framework, facilitated by GEE support, serve as potent tools for the assessment of pollutant concentration in inland water bodies on a large scale.
Recently, the number of traffic accident deaths caused by unsafe driving has been increasing dramatically. At present, unsafe driving behavior detection in urban traffic remains a challenging task due to complex background, low spatial resolution, and poor illumination conditions etc., which limit generalization ability and low detection accuracy. In this paper, a novel detection network, called UDBNet, is proposed for unsafe driving behavior detection in complex traffic monitoring scenes. UDBNet introduces a deformable global convolution block to overcome the shortcomings of the existing Convolutional Neural Network (CNN) in capturing global features and adapting to deformed targets. In addition, UDBNet adopts a new Feature Pyramid Network (FPN), termed MBFPN, to enhance the ability of feature fusion. Finally, the effectiveness, robustness, and generalization of UDBNet are proved through extensive experiments.
Road collapse has become one of the frequent geological disasters in large cities. Scientific risk assessment is of great significance for the layout of urban road collapse prevention and control work. The current standards related to road collapse risk assessment can be roughly divided into two categories: industry standard Standard for Comprehensive Detection and Risk Evaluation of Underground Disasters in Urban Area (JGJ/T 437-2018) and regional standards in Beijing City Code for Assessment and Prevention and Cure of Surrounding Soil Disease of Underground Pipeline (DB11/T 1347-2016) and Henan Province Technical Standard for Detection of Underground Disasters on Urban Roads (DBJ41/T 233-2020). For the visual prevention and control of road collapse risk, the risk assessment system based on geographic information system (GIS) has been gradually promoted. Definitely, this paper constructs a GIS-based urban road collapse risk assessment system (RCRAS) based on JGJ/T 437-2018, DB11/T 1347-2016 and DBJ41/T 233-2020. The system has developed functional modules such as ground and underground 2/3-Dimensional visualization, database management and road collapse risk assessment. Subsequently, it was applied to the road collapse risk assessment of Dongfeng Road, Guangzhou, P.R. China, which provides scientific support for the visualization and standardized implementation of urban road collapse risk prevention and control for Guangzhou Transportation Bureau.
Remote sensing image captioning (RSIC) has inspired spread attention because of its ability to flexibly generate sentences from input images. However, existing methods still have limitations in describing complex, multi-scale remote sensing image scenes. To solve this problem, we propose a multi-scale feature fusion network (MS-Net). Firstly, the features of different layers are fused to extract multi-scale feature. Further, we use the transformer encoder adaptively to fuse image features at different scales. Extensive experiments are conducted on benchmark dataset, namely NWPU-captions to verify the performance of MS-Net.
Urban perception research is essential due to the rising demand for high-quality urban spaces, but current research often overlooks the spatial heterogeneity and functional characteristics of urban environments. This study aims efficiently evaluate urban environment perception based on street view images, and couple urban spatial structure with functional characteristics to understand impact mechanisms. We associate street scene images with visual elements to evaluate urban environment perception across six dimensions, with a case in Changchun, China. Various index factors of functional characteristics are introduced to explore the influence mechanism of urban environmental perception. This research enriches the theoretical framework of urban environmental perception and provides technical support for comprehensively evaluating urban spatial quality, considering influencing factors related to the physical environment and social functions.
As an important water conservancy facility, sluice plays an important role in water conveyance, flood control, drainage and navigation. At present, the management of the sluice station is mainly based on manual control and general two-dimensional information management system. However, most sluice stations are long project lines and lots of engineering points, which easily leads to many management problems such as high monitoring and maintenance costs and difficult policy coordination. To address this problem, this study combined BIM with GIS to develop an intelligent early warning and dispatching system. Based on the BIM model, three-dimensional (3D) city building model and remote sensing data, the proposed smart sluice station system integrated GIS, BIM and IoT technology, built a smart sluice station cloud platform, and realized the remote unified management system for sluice station. As a study case, the proposed system was applied to Cijiang Sluice Station, and the results show that the proposed system can realize remote monitoring, early warning and real-time scheduling of the sluice station, which has greatly saved time and costs for the sluice management.
Cropland conversion disrupt local agricultural production systems and pose a serious threat to global food security. The use of remote sensing for change detection (CD) can detect and prevent such events in a timely manner. However, existing CD methods struggle to produce change detection results efficiently and accurately. Additionally, the limited receptive field of the convolution process prevents CNN-based approaches from catching long-range relationships. The Vision Transformer, on the other hand, excels in a variety of vision-related tasks, including picture classification, object detection, and semantic segmentation, and has significant promise for modeling long-range relationships. Because of this, in this research we suggest a UNet-like network with union attention for detecting changes in cropland using aerial remote sensing images. We utilize a Swin Transformer backbone as the encoder and an effective union-attention Transformer block to build the decoder in a Transformer-based encoder-decoder structure. The application of a multibranch prediction head with two CNN classifiers yields change maps and enhances deep layer supervision. The effectiveness and advantages of the SwinUCDNet have been demonstrated through comparative experiments with several CD methods.
Jiangsu Province in China is located along the coast and has both a temperate monsoon climate and a subtropical monsoon climate, with abundant precipitation and prone to extreme precipitation. Based on the temperature and precipitation data of 71 meteorological stations in Jiangsu Province and the European Center for Medium-Range Weather Forecasts (ECMWF) ERA5 data set, the temporal and spatial patterns of precipitation distribution in Jiangsu Province from 2001 to 2020 and the spatial changes of the impact of different factors on precipitation were analyzed by trend analysis, Moran index, and multi-scale geographically weighted regression. The results show that the precipitation in the northern part of Jiangsu Province is decreasing, while the precipitation in the southern part is increasing from 2001 to 2020. The high precipitation area is mainly concentrated in the southeast, while the low precipitation area is mainly concentrated in the west. Summer precipitation is negatively correlated with water surface pressure, and positively correlated with temperature, total evaporation, wind speed along the latitudinal line of 10 meters, wind speed along the meridional line of 10 meters, and altitude.
A large number of heterogeneous sensors are deployed in the geographic space. Observations are carried out all the time, forming massive time-series observation capability data. Observation capability is the most remarkable spatio-temporal feature of observation, which link the geographic phenomena in the real world with sensors, observation results, and sampling space-time to support the cognition of observation process and the retrieval, interpretation, and use of observation results. In this paper, the process of geospatial observation is analyzed and abstracted, and the data structure of observation capability spatio-temporal data is defined. On this basis, an observation capability spatio-temporal data model is proposed to efficiently manage multi-source, multi-dimensional, and long-time series observation capability spatio-temporal data.
The disparity of resource access (e.g., food and healthcare) among different population groups essentially reflects social inequalities. Emerging information and communications technology (ICT) has facilitated the teleactivities that can replace or complement traditional physical visits, yet existing approaches for measuring access disparity still fail to consider virtual interactions. To this end, this study proposes a unified framework to measure access inequality in both physical and virtual spaces simultaneously, using the POIs’ visit patterns from mobile phone data to capture spatial unevenness of accessibility among different groups (physical space), as well as the Household Pulse Survey data to reveal the group disparity of teleactivity access (virtual space). In particular, a novel Access Inequity Index is proposed based on the Information Theory Index and the Theil Index, to reveal the access inequality in physical and virtual spaces respectively. Next, to demonstrate the feasibility of the proposed framework, we examine racial groups and their disparity in physical-virtual healthcare access during the pandemic (April-July 2021) in U.S. 15 most populated metropolitans. Our results indicate: (1) race is a significant risk marker for underlying conditions that affect health, including telehealth access; (2) the usage of telehealth access aligns with the risk for COVID-19 infection, hospitalization, and death by race (e.g., the minority groups Black and Others are more vulnerable and in the higher demand of telehealth service); and (3) residential segregation impacts on the segregated pattern of physical healthcare access by race (e.g., the Black-dominant healthcare access zone highly matches the Black residential cluster in the south of Chicago), while such impacts may differ in different kinds of healthcare services (e.g., physicians, mental health practitioners, and dentists). Compared with traditional single-space approaches, the proposed hybrid-spaces approach not only provides more comprehensive and in-depth insights to understand the racial disparity in healthcare access, but also brings new opportunities to a broad scope of studies in social inequality measurement in future.
Based on the severe rainstorm and flood disaster in Zhengzhou city on July 20, 2020, we proposed a GIS-based urban rainstorm and flood disaster simulation analysis method, by using rainfall time series data and remote sensing image data. The results show that: (1) At 15 pm on the 20th, the rain gradually intensified with the center began to move from northwest to southeast. At 17 pm, the intensity of extraordinary rainstorm reached the maximum (187.1 mm), and waterlogging depth in some areas exceeded 40 cm. (2) Considering the actual situation of Zhengzhou City, we optimized the drainage factor of the runoff-inundation model by distinguishing 8 runoff coefficients and 22 drainage weight coefficients of the road network. (3) We compared the calculation results of the storm water inundation extrapolation model with the interpretation results of SAR images. The accuracy of the model is about 90%, which indicates that it has high accuracy and reliability in urban flood simulation, and can achieve the purpose of near real-time dynamic monitoring of flood spreading process in Zhongyuan District. This study contributes to urban emergency response and scientific decision-making, and has important significance for monitoring and evaluating the social and economic impact of flood disaster.
In order to provide a personalised experience to customers, it’s essential for shopping centers to understand its customer base and their shopping behaviors. Building a well-developed customer profile is critical for improving marketing efficiency, expanding market share, and building long-term, stable business ties with trading partners. Currently most shopping malls or retail business use footfall or customer surveys to grasp the customer behaviors, which are insufficient to obtain accurate and representative information about the customers. This study aims to provide a detailed customer profile for shopping centers using GPS datasets. We choose the two Westfield shopping malls in London as the case study area. In order to uncover additional customer information, this study focuses four research questions:(1) Origin places of customers; (2) Their transportation mode to the mall; (3) The average dwell time of customers; (4) The pattern of return visitors. According to the results, malls can develop a range of marketing initiatives to provide a better shopping experience for customers and attract more of them.
Traffic flow is the main carrier and manifestation of the flow of elements such as people and goods, and it can enrich the understanding of urban network and spatial structure. Based on the AutoNavi travel big data of 2020, this study uses methods such as data modeling, spatial analysis, and complex network analysis to analyse the spatial connection of the Guangdong-Hong Kong-Macao Greater Bay Area (GBA). The results have shown that the interurban connection intensity on the east coast of the GBA is significantly greater than that on the west coast, and cross-city commuting is mainly concentrated in the Guangzhou, Shenzhen and Hong Kong metropolitan circles. The urban network centrality has the characteristics of polarization, and the centrality of Guangzhou, Shenzhen, Foshan and Dongguan is higher than other cities. The diversity of urban network is quite different, and the diversity of Shenzhen, Macau and Hong Kong is low due to the influence of traffic location and geographical location. Generally speaking, the GBA has not yet formed a wide-ranging communication chain. It is recommended to strengthen the construction of the transportation on the west coast, promote the linkage between the East and West Coasts, and provide support for the integrated development of the GBA.
As one of the serious natural disasters, forest fire has the characteristics of strong suddenness, hard to detect, wide range of disaster, difficult to extinguish and heavy losses. Frequent forest fires put forward an urgent need for dynamic monitoring of forest fire diffusion and prediction of forest fire behavior. This paper constructs a forest fire diffusion model based on cellular automata, and outputs the visualization results of forest fire diffusion by inputting environmental factors such as terrain data, meteorological data and vegetation type data and the location of the fire point. Finally, the model is used to simulate the forest fire diffusion process in Xichang City, Sichuan Province, and the validity of the model is verified by remote sensing images. The results show that the over-fire area obtained by the simulation experiment is similar to the spatial distribution of the over-fire area extracted by Sentinel-2 satellite, which proves that the cellular automata model constructed in this paper has high accuracy.
In this paper, a new multi-UAV coverage path planning method is proposed for the scenario that UAVs need to completely search and cover a large area, which can plan an optimal feasible path that can avoid obstacles for multi-UAVs to cover the 3D space cooperatively. The method firstly decomposes the target space into mutually independent free cells by the cell decomposition algorithm; then each cell is covered individually according to the field of view of the UAV-mounted camera; then the task allocation among UAVs and the order of connecting cells are determined by comparing two classical task allocation algorithms, simulated annealing algorithm and genetic algorithm, with the total length of the path and the equilibrium degree among UAVs as the objective function; finally the task allocation among UAVs and the order of connecting cells are obtained by interpolation. Finally, the three-dimensional coverage path is obtained by interpolation. The experimental results show that the simulated annealing algorithm converges faster and with higher convergence accuracy than the genetic algorithm, and the path obtained by this method has the advantages of not being affected by terrain, multi-UAV collaboration and being able to avoid obstacles.
The dense urban metro network plays an important role in urban transportation, and it is becoming increasingly important to improve the operation and management of metro facilities. The monitoring and management of passenger flow is a main concern in metro operation, and the reliable analysis of passenger flow can greatly improve the operational efficiency and safety of a metro station. Therefore, using the long-time in-and-out smart card data of Shenzhen metro stations, this paper proposes a Coarse-to-Fine passenger flow analysis method for the characterization of passenger flow on multiple time scales. This method, which is proposed from a new perspective based on the time series clustering of metro stations, precisely defines the peak travel hours and then extracts features for the evaluation of the crowdedness and disorderliness at each station. Finally, the metro stations are classified into nine levels according to those features. The stations that need urgent attention in terms of their passenger flow, including Shenzhen North Station, Buji, and Grand Theater, are identified for the reference of city managers.
Jiaozuo City lies in the Yellow River Basin with a long history and rich spiritual cultivation. Based on geographic information system, this article analyzes the spatial distribution characteristics of historic sites in Jiaozuo City. The Yellow River plays an essential role in the creation, development, and dissemination of culture. At the same time, the terrain also impacts the spread of culture.
Under the background of new fundamental surveying and mapping construction, it is necessary to update and produce new geospatial data. Knowledge graph is an important branch technology of AI and the basis for data interconnection to knowledge interconnection services. Firstly, this paper carries out ontology modeling of geographic information. Secondly, Neo4j is used to semantically store the classification of fundamental geographic information and its geometric mapping relationship to construct the fundamental geographic information knowledge graph. Finally, experiments are conducted to verify the feasibility of graph-driven method by taking housing entity's production as an example. This method provides a new idea for the transformation and production of geospatial data in new fundamental surveying and mapping.
Weakly supervised semantic segmentation (WSSS) can well solve the problem of insufficient samples for semantic segmentation of remote sensing images. In order to better solve the problem of insufficient training samples, we introduce a Soft-MultI-Label-guIded wEakly Supervised semantic segmentation framework (SMILIES). It can generate pixel-level labels relatively under the supervision of few-shot image tag levels. We test it on the GID dataset (150 images) after training on only five images, and obtain 58.7% mIoU, which is higher than other WSSS methods when using few shot training samples. When we adopt image tag as supervision apply inferring on test data, our method has a better performance than fully-supervised DeepLabV3 with the same training samples. It can be inferred from the experiment that the SMILIES has better generalization performance and manual pixel-level labeling can benefit from it.
