Web texts contain a great deal of implicit geospatial information, which provide great potential for the geographic knowledge acquisition and service. Geographic knowledge graph is the key to extend traditional geographic information service to geographic knowledge service, and also the ultimate goal of the collection and processing of implicit geographic information from web texts. This paper systematically reviews the state of the arts of the researches on open geographic semantic web, geographic entity and relation extraction, geographic semantic web alignment, and knowledge graph storage methods. The pressing key scientific issues are also addressed, including the quality evaluation of geospatial information collected from web texts, geographic semantic understanding, spatial semantic computing model, and heterogeneous geographic semantic web alignment.
Data Model is the basis for the effective management, sharing and application of scientific data. Nowadays, the sematic data model is a conventional and dominant data organization method in the solar-terrestrial physics domain which aims to describe data along with its various metadata, such as observatories, instruments, and data types etc. However, it’s difficult to support mass data processing and correlation analysis because the model neglects the temporal and spatial relations among data. Hence, a data model supporting spatio-temporal computation should be established to facilitate data discovery, fine structure identification, coupling relation research and spatio-temporal evolution analysis and other research hotspots of solar-terrestrial physics. Therefore, this paper proposed a computable spatio-temporal data model, HTM-ST that supports these applications. On the basis of the HTM global discrete grid, this model established discrete spatio-temporal subdivision by extending HTM’s spherical units to equal-divided time dimension. Besides, a novel spatio-temporal coupled coding algorithm is described to represent these high-dimensional units in the one-dimensional space. Meanwhile, the model’s storage scheme is designed and implemented in the HBase platform based on the model’s subdivision structure and coding algorithm. In this paper, a prototype system is implemented to evaluate the efficiency of the model, by comparing multiple spatio-temporal queries over energetic particle data observed by five polar orbit satellites. The experimental results show that HTM-ST data model is more efficient and robust. It could be used as the solar-terrestrial physics data organization and storage foundation for spatio-temporal relationship.
Currently, the localization of users, such as people, vehicles or robots, in indoor spaces is a common issue in many commercial and industrial applications. A number of technologies have been proposed for indoor localization based on different principles such as RF (Radio Frequency), magnetic fields, ultra wide band (UWB) and ultrasound. Among these up-to-date indoor positioning technologies, most of them depend on special infrastructures or devices, which limit the commercial application of indoor localization. As one of the state-of-art indoor localization method, visual-based positioning scheme do not rely on any external auxiliary equipment and consequently has the advantage of low cost. However, the construction of geo-tagged image database, one of the most important parts for visual-based localization, is quite labor-intensive and time consuming. The automatic collection of geo-tagged indoor image data is an essential bottleneck for application of visual-based indoor localization systems. This paper proposed a visual-based indoor positioning approach which can automatically collect geo-tagged images based on the integration of structure from motion (SFM) and pedestrian dead rocking (PDR). The main idea of this method is to collect video frames as well as inertial data (by using smartphones) when people are walking in indoor environments. A method is designed to estimate the location (i.e., geo-tags) of images for the construction of geo-tagged image database. There are two phases for this approach: offline phase and online phase. During offline phase, the proposed method is used to estimate the location of the images extracted from video frames. A multi-constrain image matching algorithm was also developed to improve the performance of location estimation. There are three constraints in this multi-constrain image matching algorithm: ratio constraint, symmetry constraint and RANSAC constraint. Based on this image matching algorithm, a SFM process can be conducted to estimate the heading angle of a walking trajectory. After that, the coordinate of sampling points from the walking trajectory can be estimated by using the PDR method and the geo-tagged image database can be constructed. During the online phase, an indoor localization method is proposed to estimate the location of a pedestrian by finding the best matching images of a query image (taken by the pedestrian) from the image database. The multi-constraint image matching algorithm can be used to compute the number of matched key-points between query images and database images. The images with the most matched key-points are selected as the candidates. A weighted average function is used to estimate the location of the query image based on the selected images. In order to evaluate the performance of the visual-based indoor localization approach, two types of indoor environment are selected as the study area: an office building and a hospital. The experimental result showed that the average location estimation error of the geo-tagged images was 0.58 m. Based on the constructed image database, the error of the proposed localization approach ranged from 0.2 to 1.4 m. The performance of this approach indicated that visual information contribute to the construction of geo-tagged image database. This approach can be used in various indoor environments without any infrastructures or extra devices, which can reduce the difficulty in applying this approach to practical use.
As one of the most typical artificial landforms in loess plateau of China, artificial terraces have important value in agricultural production as well as soil and water conservation. The previous studies on terrace automatic extraction only focused on their boundaries and failed to achieve effective and automatic extraction of their ridge lines. In this study, a quick extraction method of terraces based on illumination and shading simulation of DEM is proposed. Firstly, illumination and shading simulation in four-direction is carried out using 1 m digital elevation model (DEM) generated by Unmanned Aerial Vehicle (UAV) technology and average value of four simulations was calculated. Secondly, the image of the mean value is binarized into a logistic image using a proper threshold. Also, valley and other non-terraced areas were masked off. Finally, terrace ridge lines were obtained by automatic vectorization of the logistic image. Some broken lines were filtered by a proper length threshold to improve the extraction precision. Taken Wangdonggou watershed, Changwu County, Shaanxi Province as the study area, the experiment results showed that this method had a preferable extraction accuracy of 89.09%. Furthermore, the parameters involved in this method were also discussed. The results showed that if we take two orthogonal symmetric direction angles as the direction angles, arctangent value of the slope of terrace ridges as the altitude angle, an empirical formula (t=180-σ) was used to calculate the threshold value of the logistic image, the quick automatic extraction of artificial terraces of loess plateau can be achieved.
Call detail records （CDRs） have been widely used to study human activities over the world. They offer a new channel to estimate population distribution with higher spatiotemporal resolution. However, the samples of CDRs distribute irregularly and sparsely, which could cause certain bias in the derived population distribution. This study is the first assessment that takes a mobile signaling dataset of 24-hour tracking users as a benchmark to evaluate the bias in population distribution derived from CDRs. Particularly, taking Shenzhen City as an example, this study quantifies the relative errors of 24-hour population distributions from both temporal and geographical dimensions, and also discusses the impact of excluding low-frequency call users on these errors. This study found that the medians of relative errors lie between 25%~30% when using caller volume to estimate population distribution during human active hours and the errors will increase during sleeping time. Such bias should be made aware of for researchers and application practitioners. This study also demonstrated that the urban land use types strongly relate with estimation errors of population distribution derived from CDRs. Especially, the population in rural residential land and industrial land will be significantly underestimated, while that in transportation land will be highly overestimated. For applications such as emergency evacuation or facility allocation based on population derived from CDRs, these results can support correcting population estimation errors and help to locate rescuing support or public resources more properly. At last, this study showed that excluding low-frequency call users can slightly mitigate the impact of land use on the estimation errors, suggesting excluding low-frequency users in a rigorous way. Overall, the findings of this study can help understand the limitation and suitability of applying CDRs to estimate population distribution with high spatiotemporal resolution, as well as offering scientific support for research and applications of using CDRs in an appropriate way.
The prediction of the service population of cell phone tower plays an important role in the optimization of the spatial location of the cell phone towers and the configuration of the communication service bandwidth in mobile communications, and also provides the decision-making basis for early warning of human convergence and prevention of group incidents in urban management. This study proposed a prediction method of population in a region on a citywide scale, considering the human movement between cell phone towers. Based on the Markov chain and Bayesian probability theorem, we calculated the transition probability between different cell phone towers using massive mobile phone trajectories and we acquired the transition probability matrix which was distinct in different periods of one day. We made full use of the spatiotemporal transition probability to quantify the spatial and temporal characteristics of the mobile phone user’s intra-urban movement. This study applies massive historical mobile phone location data to model training and proposes a citywide prediction model of the service population of the cell phone tower based on the spatiotemporal transition probability model. The algorithm proposed is verified by a mobile phone location dataset within thirty days collected by Hubei Mobile. This study shows that the prediction accuracy rate can reach about 94.8% and the proposed algorithm performed a good prediction with a temporal granularity of 60 minutes. Moreover, this study analyzed the prediction performance of the spatiotemporal transition probability model in different time granularities, and made comparison with other methods, such as the Castro model and moving average method. The results indicated that the proposed algorithm outperforms the Castro’s model and moving average method when the temporal granularity is larger than 20 minutes. The proposed prediction method takes account of the spatial and temporal characteristics of human mobility and provides more accurate prediction results.
Spatial clustering is one of the most important methods in spatial data mining. As a common but powerful spatial clustering algorithm, K-Medoids is applied in many fields such as generalization of spatial entity information, spatial point pattern analysis and epidemiology application. However, K-Medoids algorithm meets two main challenges innately as follow. At first, K-Medoids has selection problem of the initial medoids. Different initial medoids may not attain the same clustering results which could lead to a non-optimal results sometimes. Furthermore, time efficiency of the algorithm is not satisfactory because there exist quantities of iterations to find the most suitable partition. Existing studies on the K-Medoids algorithm don’t take the validness and time efficiency into consideration at the same time. Optimal methods like the Genetic Algorithm are applied to improve the validness of K-Medoids but the time efficiency is not acceptable when dealing with growing data. The MapReduce model is utilized to handle with data of high volume which can’t adapt to some circumstances short of computer clusters. In order to improve the result validity and time efficiency of the algorithm, this paper revised the traditional K-Medoids algorithm of Partitioning Around Medoids (PAM) combining with the idea of the Simulate Anneal Arithmetic (SAA) and proposed a parallel Simulate Anneal Partitioning Around Medoids (SAPAM) algorithm which was implemented efficiently in Graphics Processing Units (GPUs). SAA algorithm is used to search for the initial medoids which promises the validness of the algorithm. The stochastic factor introduced in SAA algorithm gives the possibility of eliminating the local optima to attain the global optimal clustering results of PAM. To accelerate the clustering process, we design the parallel SAPAM algorithm to utilize quantities of GPU’s threads which execute the program at the same time. By analogy with the matrix multiplication, a new matrix computation method is defined to reduce the time consumption of data transfer between GPU’s global memory and shared memory. The matrix computation method reuses data in the shared memory of GPU and computes the distances between medoids and many points at a time which improve the algorithm’s performance evidently. To validate the proposed algorithm, we generated eight datasets with different attributes and sizes randomly and conducted experiments on the eight datasets to compare the proposed parallel SAPAM algorithm with the traditional PAM algorithm, sequential SAPAM algorithm and the parallel genetic K-Medoids algorithm. The experiment results showed that SAPAM algorithm attained more accurate clustering results compared with the traditional PAM and the parallel genetic K-Medoids algorithm. Besides, the proposed algorithm performed better than the sequential SAPAM algorithm and the parallel genetic K-Medoids algorithm in time efficiency. According to the results, our GPU-based SAPAM algorithm was four to eight times faster than the traditional PAM algorithm. The results demonstrate that the proposed method can execute efficiently and attain a valid result. Finally, SAPAM algorithm was applied to analyze the safety monitoring data of Guizhou province to get the clustering pattern of the safety threats. The clustering results show us several clusters of the safety threats which may provide some practical application value to the governor.
Analysis of multiple time series land use of spatio-temporal evolution at fine scale is a hot and important research area currently. In this study, the self-organizing map （SOM） neural network was used to analyze the land use spatio-temporal distribution at township-level in Beijing. The study was based on 5 periods of land use classification data of Beijing in 2005, 2007, 2009, 2011 and 2013. We implemented spatio-temporal integrated expression and comparative analysis of multiple time series land use data at township-level. Through creating and training self-organizing map neural network, we could find out the distribution of different land use types （built-up land, farmland, forest land, grassland, garden, water, and unused land） on the SOM output plane. This represented the proportional relationship of different land use types in land use structure. By second-step clustering and building land use change trajectory, we got the spatio-temporal evolution rules of the land use in township of Beijing. The results revealed that there were five land use change trajectories and three spatio-temporal evolution patterns in Beijing at township level. The plain area is developing to the land use structure of high built-up land proportion. The mountainous area is developing to the land use structure of high forest land proportion, and the land use change of piedmont zone is complex.
Urban population distribution and its dynamic changes have been playing a key role in urban planning and management. Currently, the wide use of information communication technology （ICT） provides an opportunity to support fine-scale studies by acquiring accurate individual positioning data. By extracting regular individual trajectories from a mobile communication signaling dataset, this study established an estimation procedure of urban population distribution and quantitatively analyzed spatiotemporal characteristics of population distribution and migration in Shanghai. The results indicated that, firstly, mobile communication signaling data had the ability to describe the dynamic characteristics of urban population and to estimate the real population size of a city in a quantitative and relatively an authentic way by taking its advantages of wide sample coverage, high spatial resolution, good timeliness and multiple spatiotemporal scales. Secondly, population distribution of Shanghai on the whole is stable all day long. Comparatively, population at the daytime showed a more remarkable spatial agglomeration phenomenon than population at night. Thirdly, the population migration between urban functional areas and other areas is rare. During the rush hour in the morning and evening, the population migration was mainly depicted as a relatively equally both-way movements between central urban area and other new urban functional areas. Within each functional area, more than half of its population is not moving out. In conclusion, this study can be useful for urban planning, emergency management and public traveling information services.
Based on the summary of the diversified modes, multi-level traffic nodes, efficient operation and structure, the evaluation model of the flow-carrying capacity of the urban agglomeration traffic network is constructed with GIS platform. Take Shandong Peninsula urban agglomeration （SPUA） as a case, this study aims to evaluate and optimize the comprehensive transportation network in urban agglomerations with traffic network flow-carrying model （TNFM）. First, the contrast analytical matrix was constructed, in which the practical flow-carrying values for traffic lines can be compared with their theoretical values. Second, using the spatial analysis tool of GIS, the performances of comprehensive traffic network as well as every type of traffic lines were evaluated. The national road, provincial road and railway of the urban agglomeration is inadequate. Diversion of some urban trunk road is poor. 60% of the administrative units at all levels have shortage of road. High-speed rail and county road can meet the current carrying requirements. In the future, SPUA should focus on strengthening the construction of the above four types of traffic lines, improve the traffic interface system, and give full diversion to the role of high-speed rail and county road.
Geographic Object-Based Image Analysis (GEOBIA) is much better than traditional pixel-based method of high spatial resolution remote sensing image analysis. Since image segmentation is the key technique in GEOBIA, scholars and researchers have already conducted extensive research and proposed a number of segmentation algorithms. In order to compare different segmentation methods and evaluate its own performance, segmentation results need to be evaluated. Therefore, the study of segmentation evaluation is equally important to segmentation algorithm. We could choose the applicable segmentation method and set appropriate parameters for specific images and applied the segmentation evaluation. The aim of image segmentation is to enable the automation of image analysis. However, the evaluation methods which cannot provide quantitative indexes are not applicable in automatic real-time image analysis system. Moreover, research in segmentation evaluation is less than segmentation itself. Thus, it will be significant to study segmentation and review the quantitative evaluation method. In this paper, based on summarizing the evaluation methods, the hierarchy of segmentation evaluation method is presented. In spite of describing quantitative empirical methods, we discussed their range of application. Their advantages and shortcomings were also analyzed. Finally, possible future direction and potential application prospect for high spatial remote sensing image segmentation evaluation were proposed.
As the space for human habitation and activity, urban buildings are an important part of the city. Their renewal and renovation affects development of the city and changes people’s life at all times. Functional classification of urban buildings provides supporting evidence for categorizing urban functional areas, and also helps the government in land use planning, as well as managing the distribution of population and resources, promoting the sustainable development of urban areas. However, current classification technology of remote sensing is insufficient to make functional classification of urban buildings. In this paper, we analyzed urban information in great depth, by classifying the function of urban buildings. The efficiency and precision of the classification is improved after combining remote sensing, the Internet POI (Point of Interest) data and GIS technology. We first chose the method of CNN (Convolutional Neural Networks) to extract building information from remote sensing images of high resolution. The precision of the extraction is above 93% as is shown by precision evaluation. POI data was then classified into 3 types by manual work, namely buildings used for commercial service, public service and residence. The classified POI data were estimated by Kernel Density. After which the mean Kernel Density value of every type of buildings was calculated and these three types of buildings were delimited by thresholds. Thus, buildings for commercial service, public service and residence could be recognized from the building information assisted by POI data, achieving the functional classification of urban buildings. This method has shown good extraction efficiency compared to visual interpretation-the overall accuracy is 86.85% and Kappa Coefficient is 0.8153 according to precision evaluation. In future research, this method can be used to classify and identify different types of urban buildings. However, there are still some problems to be discussed in this method. For example, when defining buildings’ functional types by threshold of Kernel Density, one building may have more than one or no type. Besides, POI data have some limitations when representing the range of different types of buildings: one point may represent either a grand shopping mall or a convenience store. These will be addressed in future studies.
By identifying the research gap in the quantitative study of the visual quality of street landscape, a new Green View Index is proposed to assess the human-viewed street. The newly developed method describes the green-appearance area of street greenery that can be captured and its spatial distribution via simulating the real scene of the surrounding trees in the human eye field. Results indicate that the proposed Green View Index can reflect the impact of the size of tree canopy, layout of buildings, and distances between trees and viewers on the visual perception of street greenery in complex road scenes. The advantages of this method are as follows: by making full use of multi-source information, the green view of any location can be calculated with high precision based on automatic interpretation, which greatly reduces manual effort and errors and facilitates multi-regional comparative evaluation. The effectiveness of multi-source remotely sensed data in evaluating the visual effect has been proved, and the new Green View Index can be a relatively objective measurement in guiding urban landscape planning and management.
At present, there are several problems in mapping winter wheat using remote sensing technology at regional scale. These problems can be the differences in phenology of winter wheat, complex ground environment and data-processing, redundant remotely sensed data, difficulty of choosing appropriate samples and low accuracy. In order to solve these problems, a novel method was proposed and tested in Henan province. 2296 scenes of Landsat images in 2002 and 2015 were processed using Google Earth Engine. Google Earth Engine is the most advanced cloud-based geospatial processing platform in the world. It combines Google-scale storage and processing power in order to make substantial progress on global challenges involving large geospatial datasets. A novel method called Normalized Difference Vegetation Index (NDVI)-remodel-amplification was proposed to construct a universal model for mapping winter wheat at regional scale. The steps of the method is as follows: Landsat images from September 15 to November 15 were chosen to compute NDVI. Then, we selected the minimum NDVI as the first sequence of NDVI (recorded as NDVI1) at the pixel scale. In the same way, Landsat images from December 1st to March 31st were chosen to compute NDVI. Then, we selected the maximum NDVI as the second sequence NDVI (recorded as NDVI2) at the pixel scale. Then, amplification between NDVI1 and NDVI2 was computed and recorded as NDVIincrease. A pixel would be regarded as winter wheat if its NDVIincrease value is more than 1.3 and its NDVI2 value is more than 0.34. The results showed that winter wheat is mainly located in the middle-eastern plains and in Nanyang basin of Henan province with the characteristics of concentrated and contiguous distribution. The planting area of winter wheat in 2015 and 2002 was 56 055.79 km2 and 47 296.11 km2, respectively, with an accuracy of 97% based on statistical data. From 2002 to 2015, there was a significant change in the distribution of winter wheat in Henan Province The trend of overall sown area was increasing. Compared with that in 2002, the area of winter wheat in 2015 increased by 8759.69 km2 or 18.52%. Comparing with conditional winter wheat mapping method, this proposed method is based on Google Earth Engine showing a great improvement in both of data-processing and mapping efficiency.
Soil moisture is one of the core variables in land surface ecosystem and energy cycle. For the strong penetration of the cloud, rain and atmosphere, microwave remote sensing has advantages of the high precision in soil moisture retrieval. Currently, there are many passive microwave sensors or satellites used for surface soil (<5 cm) moisture observations, such as NASA's AMSR-E (The Advanced Microwave Scanning Radiometer-Earth Observing System) and SMAP (the Soil Moisture Passive and Active) and the European Space Agency SMOS (The Soil Moisture and Ocean Salinity). Although the use of microwave sensor can get higher precision of soil moisture products. The errors of SMAP 36 km soil moisture products can be less than 0.04 m3/m3. The 2~3 days revisited time restricts the applications that need the soil moisture products with higher temporal resolution (1 days). Therefore, it has been drawn more and more attention to get the accurate soil moisture with higher temporal resolution for the global weather prediction. Although the SM retrieval from MODIS data has higher error than retrieval from passive microwave data, the temporal resolution of MODIS data (1 day) is higher than the passive microwave data. For the different advantages of MODIS and passive microwave data, the combination of the two data for soil moisture retrieval may get the SM products with the MODIS temporal resolution and the similar accuracy or similar spatial variation of passive microwave data. In this study, we attempt to combine SMAP 36 km soil moisture product and MODIS optical/thermal infrared data to estimate the global 36 km soil moisture. This improve the temporal resolution of SMAP soil moisture from the 2~3 days to 1 day. By using the generalized regression neural network (GRNN) method, we simulated the relationship of SMAP soil moisture with MODIS global surface temperature and the surface reflectance products. Then we estimated the global 36 km soil moisture using the GRNN simulated relationship. In order to prevent overfitting of GRNN, all sample data according to the ratio of 0.8:0.2 is divided into training dataset and validation dataset. With the increase of the spread factor, the performance of GRNN prediction of the validation dataset shows a decreasing trend after the first increase, and GRNN obtained the maximal correlation coefficient (r) and root mean square error (rmse) with 0.02 of the diffusion factor. Finally, the well trained GRNN is used to estimate the global 36 km soil moisture. The results show that the accuracy of the GRNN for soil moisture estimate has a high correlation with SMAP (r=0.7528), but it retains a high error (RMSE=0.0914 m3/m3). For the cloud contamination of MODIS data, there has a part of loss of GRNN 36 km soil moisture estimate. Nevertheless, the GRNN estimated soil moisture can be very good to maintain the overall spatial variation of SMAP soil moisture, and enhance the temporary resolution of soil moisture from 2~3 days to one day. Besides, the relationship between SMAP and MODIS data is also studied in this paper, which can provide a significant reference for SMAP 36 km soil moisture downscaling by the machine learning.
