ISPRS Int. J. Geo-Inf., Volume 11, Issue 4 (April 2022) – 58 articles

Quantitative relations between topological similarity degree and map scale change of multi-scale contour clusters are vital to the automation of map generalization. However, no method has been proposed to calculate the relations. This paper aims at filling the gap by proposing a new approach. It firstly constructed a directed contour tree by pre-processing of unclosed contours, and then developed a quantitative expression of topological relations of contour cluster based on directed contour tree. After this, it employed 108 groups of multi-scale contour clusters with different geomorphological types to explore the changing regularity of topological indices with map scale. Last, it used 416 points to calculate the quantitative relations between topological similarity degree and map scale change by curve fitting method. The results show that the quantitative expression of multi-scale topological indexes is closely related to the contour interval change, and power function is the best fit among the candidate functions. Full article

Geographical information is gaining new momentum as an analysis and visualization tool for collections of cultural objects. It provides all kinds of users with new opportunities to contextualize and understand these objects in ways that resemble our ordinary spatially-located experience and to do so better than textual narratives. The SeMap project has built an online resource that shows more than 200,000 cultural objects through spatiotemporal maps, thus enabling new experiences and perspectives around these objects. Data come from the CER.ES repository and were created by a network of more than 100 Spanish museums. This article explains the refinement of the data provided by the repository, mostly by adding a semantic structure thanks to the CIDOC-CRM ontology, and by simplifying the exceedingly complex terminologies employed in the original records. Particular attention is paid to the methods for geolocating the information, as well as adding temporal filters (among others) to user queries. The functionalities, interface, and technical requirements are also explored at length. Full article

The cross-impact of environmental pollution among cities has been reported in more research works recently. To implement the coordinated control of environmental pollution, it is necessary to explore the structural characteristics and influencing factors of the PM_2.5 spatial correlation network from the perspective of the metropolitan area. This paper utilized the gravity model to construct the PM_2.5 spatial correlation network of ten metropolitan areas in China from 2019 to 2020. After analyzing the overall characteristics and node characteristics of each spatial correlation network based on the social network analysis (SNA) method, the quadratic assignment procedure (QAP) regression analysis method was used to explore the influence mechanism of each driving factor. Patent granted differences, as a new indicator, were also considered during the above. The results showed that: (1) In the overall network characteristics, the network density of Chengdu and the other three metropolitan areas displayed a downward trend in two years, and the network density of Wuhan and Chengdu was the lowest. The network density and network grade of Hangzhou and the other four metropolitan areas were high and stable, and the network structure of each metropolitan area was unstable. (2) From the perspective of the node characteristics, the PM_2.5 spatial correlation network all performed trends of centralization and marginalization. Beijing-Tianjin-Hebei and South Central Liaoning were “multi-core” metropolitan areas, and the other eight were “single-core” metropolitan areas. (3) The analysis results of QAP regression illustrated that the top three influencing factors of the six metropolitan areas were geographical locational relationship, the secondary industrial proportion differences, respectively, and patent granted differences, and the other metropolitan areas had no dominant influencing factors. Full article

In order to cope with the rapid growth of spatiotemporal big data, data organization models based on discrete global grid systems have developed rapidly in recent years. Due to the differences in model construction methods, grid level subdivision and coding rules, it is difficult for discrete global grid systems to integrate, share and exchange data between different models. Aiming at the problem of information fusion between a GeoSOT grid and global hexagonal equal area grid system, this paper proposes the GeoSOT equivalent aggregation model (the GEA model). We establish a spatial correlation index method between GeoSOT grids and global hexagonal equal area grids, and based on the spatial correlation index, we propose an interoperable transformation method for grid attributes information. We select the POI (points of interest) data of Beijing bus and subway stations and carry out the transformation experiment of hexagonal grid to GeoSOT grid information so as to verify the effectiveness of the GEA model. The experimental results show that when the 17th-level GeoSOT grid is selected as the particle grid to fit the hexagonal grid, the accuracy and efficiency can be well balanced. The fitting accuracy is 95.51%, and the time consumption is 30.9 ms. We establish the associated index of the GeoSOT grid and the hexagonal grid and finally realized the exchange of information. Full article

The magnitude of highly dynamic spatial data is expanding rapidly due to the instantaneous evolution of mobile technology, resulting in challenges for continuous queries. We propose a novel indexing approach model, namely, the Velocity SpatioTemporal indexing approach (VeST), for continuous queries, mainly Continuous K-nearest Neighbor (CKNN) and continuous range queries using Apache Spark. The proposed structure is based on a selective velocity partitioning method, i.e., since different objects have varying speeds, we divide the objects into two sets according to the actual mean speed we calculate before building the index and accessing data. Then the adopted indexing structure base unit comprises a nonoverlapping R-tree and a two dimension grid. The tree divides the space into nonoverlapping minimum bounding regions that point to the grids. Then, the uniform grid stores the object data of leaf nodes. This access method reduces the update cost and improves response time and query precision. In order to enhance performances for large-scale processing, we design a compact multilayer index structure on a distributed setting and propose a CKNN search algorithm for accurate results using a candidate cell identification process. We provide a comprehensive vision of our indexing model and the adopted query technique. The simulation results show that for query intervals of 100, the proposed approach is 13.59 times faster than the traditional approach, and the average time of the VeST approach is less than 0.005 for all query intervals. This proposed method improves response time and query precision. The precision of the VeST algorithm is almost equal to 100% regardless of the length of the query interval. Full article

Existing optical remote sensing image change detection (CD) methods aim to learn an appropriate discriminate decision by analyzing the feature information of bitemporal images obtained at the same place. However, the complex scenes in high-resolution (HR) remote images cause unsatisfied results, especially for some irregular and occluded objects. Although recent self-attention-driven change detection models with CNN achieve promising effects, the computational and consumed parameters costs emerge as an impassable gap for HR images. In this paper, we utilize a transformer structure replacing self-attention to learn stronger feature representations per image. In addition, concurrent vision transformer models only consider tokenizing single-dimensional image tokens, thus failing to build multi-scale long-range interactions among features. Here, we propose a hybrid multi-scale transformer module for HR remote images change detection, which fully models representation attentions at hybrid scales of each image via a fine-grained self-attention mechanism. The key idea of the hybrid transformer structure is to establish heterogeneous semantic tokens containing multiple receptive fields, thus simultaneously preserving large object and fine-grained features. For building relationships between features without embedding with token sequences from the Siamese tokenizer, we also introduced a hybrid difference transformer decoder (HDTD) layer to further strengthen multi-scale global dependencies of high-level features. Compared to capturing single-stream tokens, our HDTD layer directly focuses representing differential features without increasing exponential computational cost. Finally, we propose a cascade feature decoder (CFD) for aggregating different-dimensional upsampling features by establishing difference skip-connections. To evaluate the effectiveness of the proposed method, experiments on two HR remote sensing CD datasets are conducted. Compared to state-of-the-art methods, our Hybrid-TransCD achieved superior performance on both datasets (i.e., LEVIR-CD, SYSU-CD) with improvements of 0.75% and 1.98%, respectively. Full article

Determining how to reasonably allocate shelters in the central area of the city and improve evacuation efficiency are important issues in the field of urban disaster prevention. This paper introduces the methodology and mathematical model from the field of crowd emergency evacuation to shelter location optimization. Moreover, a shelter location optimization method based on the combination of static network analysis and dynamic evacuation simulation is proposed. The construction costs and evacuation times are taken as the objective functions. In the first stage, based on the static network analysis, a circular evacuation allocation rule based on the gravity model is proposed, and the genetic algorithm is then designed to solve the feasible schemes with the lowest shelter construction costs. In the second stage, the evacuation time is taken as the optimization objective. The age differences of refugees, the selection of evacuation routes, and the behavior of adults helping children and the elderly are simulated in a dynamic evacuation simulation model. The traditional social force model is improved to conduct a regional evacuation simulation and determine the optimal scheme with the shortest evacuation time. Finally, the central urban area of Xinyi City, Jiangsu Province, China, is taken as an empirical case. Full article

Spatial media bring out new forms of interaction with places, leading to the emergence of new ways of embodying the experience. The perception of place and its dynamics of change has been multiplied by the emergence of digital platforms, which create many and varied representations of place in spatial media. These representations are dependent on the digital platforms’ ecosystem, formed by platform-specific mechanisms of digital placemaking. The study applied text mining techniques and statistical methods to explore the role of user-generated content as a digital placemaking practice in shaping place experience. The online reviews were collected from Google Maps for 23 places from Poznań, Poland. The analysis showed that place experience is described by three dimensions: attributes, practices and atmosphere, or place practices that most closely reflect the specificity of a place. The place attributes blurred the boundaries between their digital images, whereas the atmosphere dimension reduces the diversity and uniqueness of the place. In conclusion, user-generated content (UGC) as an element of the process of digital placemaking increases place awareness and democratizes human participation in its creation, yet it affects its reduction to homogeneous information processed through mechanisms operating within a given digital platform. Full article

This paper aims to explore the spatiotemporal pattern of traffic accidents using five years of data between 2015 and 2019 for the Irbid Governorate, Jordan. The spatial pattern of traffic-accident hotspots and their temporal evolution were identified along the internal and arterial roads network in the study area using spatial autocorrelation (Global Moran I index) and local hotspot analysis (Getis–Ord Gi*) techniques within the GIS environment. The study showed a gradual increase in the reported traffic accidents of approximately 38% at the year level. The analysis of traffic accidents at the severity level showed a distinguished spatial distribution of hotspot locations. The less severe traffic accidents (~95%) occurred on the internal road network in the Irbid Governorate’s towns where the highest traffic volume exist. The spatial autocorrelation analysis and the Getis–Ord Gi* statistics with 99% of significance level showed clustering patterns of traffic accidents along the internal and the arterial road network segments. Between 2015 and 2019, a notable evolution of the traffic-accident hotspots clusters was pronounced. The results can be used to guide traffic managers and decision makers to take appropriate actions for enhancing the hotspot locations and improving their traffic safety status. Full article

This paper aims to present possible areas to plant different vegetation types near traffic jams to reduce air pollution in the capital of Croatia, the city of Zagreb. Based on main traffic road and random forest machine learning using WorldView-2 European cities data, potential areas are established. It is seen that, based on a 10 m buffer, there is a possible planting area of more than 220,000 square meters, and based on 15 m buffer, there is a possible planting area of more than 410,000 square meters. The proposed plants are Viburnum lucidum, Photinia x fraseri, Euonymus japonicus, Tilia cordata, Aesculus hippocastanum, Pinus sp., Taxus baccata, Populus alba, Quercus robur, Betula pendula, which are characteristic for urban areas in Croatia. The planting of proposed trees may result in an increase of 3–5% in the total trees in the city of Zagreb. Although similar research has been published, this paper presents novelty findings from combined machine learning methods for defining green urban areas. Additionally, this paper presents original results for this region. Full article

This editorial introduces the Special Issue entitled “Geomatic Applications to Coastal Research: Challenges and New Developments” and succinctly evaluates future trends of the use of geomatics in the field of coastal research. This Special Issue was created to emphasize the importance of using different methodologies to study the very complex and dynamic environment of the coast. The field of geomatics offers various tools and methods that are capable of capturing and understanding coastal systems at different scales (i.e., time and space). This Special Issue therefore features nine articles in which different methodologies and study cases are presented, highlighting what the field of geomatics has to offer to the field of coastal research. The featured articles use a range of methodologies, from GIS to remote sensing, as well as statistical and spatial analysis techniques, to advance the knowledge of coastal areas and improve management and future knowledge of these areas. Full article

Groundwater is one of the most important natural resources for reliable and sustainable water supplies in the world. To understand the use of water resources, the fundamental characteristics of groundwater need to be analyzed, but in many cases, in situ data measurements are not available or are incomplete. In this study, we used GIS and fuzzy analytic hierarchy process (FAHP) techniques for delineation of the groundwater potential zones (GWPZ) in the Titel Municipality (northern Serbia) based on quantitative assessment scores by experts (hydrologists, hydrogeologists, environmental and geoscientists, and agriculture experts). Six thematic layers, such as geology, geomorphology, slope, soil, land use/land cover, and drainage density were prepared and integrated into GIS software for generating the final map. The area falls into five classes: very good (25.68%), good (12.10%), moderate (15.18%), poor (41.34%), and very poor (5.70%). The GWPZ map will serve to improve the management of these natural resources to ensure future water protection and development of the agricultural sector, and the implemented method can be used in other similar natural conditions. Full article

Image recording is now ubiquitous in the fields of endangered-animal conservation and GIS. However, endangered animals are rarely seen, and, thus, only a few samples of images of them are available. In particular, the study of endangered-animal detection has a vital spatial component. We propose an adaptive, few-shot learning approach to endangered-animal detection through data augmentation by applying constraints on the mixture of foreground and background images based on species distributions. First, the pre-trained, salient network U2-Net segments the foregrounds and backgrounds of images of endangered animals. Then, the pre-trained image completion network CR-Fill is used to repair the incomplete environment. Furthermore, our approach identifies a foreground–background mixture of different images to produce multiple new image examples, using the relation network to permit a more realistic mixture of foreground and background images. It does not require further supervision, and it is easy to embed into existing networks, which learn to compensate for the uncertainties and nonstationarities of few-shot learning. Our experimental results are in excellent agreement with theoretical predictions by different evaluation metrics, and they unveil the future potential of video surveillance to address endangered-animal detection in studies of their behavior and conservation. Full article

At present, a common drawback of crowd simulation models is that they are mainly simulated in (abstract) 2D environments, which limits the simulation of crowd behaviors observed in real 3D environments. Therefore, we propose a deep reinforcement learning-based model with human-like perceptron and policy for crowd evacuation in 3D environments (HDRLM3D). In HDRLM3D, we propose a vision-like ray perceptron (VLRP) and combine it with a redesigned global (or local) perceptron (GOLP) to form a human-like perception model. We propose a double-branch feature extraction and decision network (DBFED-Net) as the policy, which can extract features and make behavioral decisions. Moreover, we validate our method’s ability to reproduce typical phenomena and behaviors through experiments in two different scenarios. In scenario I, we reproduce the bottleneck effect of crowds and verify the effectiveness and advantages of HDRLM3D by comparing it with real crowd experiments and classical methods in terms of density maps, fundamental diagrams, and evacuation times. In scenario II, we reproduce agents’ navigation and obstacle avoidance behaviors and demonstrate the advantages of HDRLM3D for crowd simulation in unknown 3D environments by comparing it with other deep reinforcement learning-based models in terms of trajectories and numbers of collisions. Full article

This paper presents an approach to GWS (GeospatialWeb Service) discovery through the semantic annotation of WPS (Web Processing Service) service descriptions. The rationale behind this work is that search engines that use appropriate semantic-based similarity measures in the matching process are more accurate in terms of precision and recall than those based on syntactic matching alone. The lack of semantics in the description of services using a standard such as WPS prevents the use of such a matching process and is considered a limitation of GWS discovery. The GWS discovery approach presented is based on the consideration of semantics in the service description method and in the matching process. The description of services is based on a semantic lightweight meta-model instantiated in the WPS 2.0 standard, extending the description of the service through metadata tags. The matching process is performed in three steps (functionality matching step, I/O (Input/Output) matching step and non-functional matching step). Its core is a semantic similarity measure that combines logical and non-logical matching methods. Finally, the paper presents the results of an experiment applying the proposed discovery approach on a GWS corpus, showing promising results and the added value of the three-step matching process. Full article

In order to realize the management of various street objects in smart cities and smart transportation, it is very important to determine their geolocation. Current positioning methods of street-view images based on mobile mapping systems (MMSs) mainly rely on depth data or image feature matching. However, auxiliary data increase the cost of data acquisition, and image features are difficult to apply to MMS data with low overlap. A positioning method based on threshold-constrained line of bearing (LOB) overcomes the above problems, but threshold selection depends on specific data and scenes and is not universal. In this paper, we propose the idea of divide–conquer based on the positioning method of LOB. The area to be calculated is adaptively divided by the driving trajectory of the MMS, which constrains the effective range of LOB and reduces the unnecessary calculation cost. This method achieves reasonable screening of the positioning results within range without introducing other auxiliary data, which improves the computing efficiency and the geographic positioning accuracy. Yincun town, Changzhou City, China, was used as the experimental area, and pole-like objects were used as research objects to test the proposed method. The results show that the 6104 pole-like objects obtained through object detection realized by deep learning are mapped as LOBs, and high-precision geographic positioning of pole-like objects is realized through region division and self-adaptive constraints (recall rate, 93%; accuracy rate, 96%). Compared with the existing positioning methods based on LOB, the positioning accuracy of the proposed method is higher, and the threshold value is self-adaptive to various road scenes. Full article

The classification of unmanned aerial vehicle hyperspectral images is of great significance in agricultural monitoring. This paper studied a fine classification method for crops based on feature transform combined with random forest (RF). Aiming at the problem of a large number of spectra and a large amount of calculation, three feature transform methods for dimensionality reduction, minimum noise fraction (MNF), independent component analysis (ICA), and principal component analysis (PCA), were studied. Then, RF was used to finely classify a variety of crops in hyperspectral images. The results showed: (1) The MNF–RF combination was the best ideal classification combination in this study. The best classification accuracies of the MNF–RF random sample set in the Longkou and Honghu areas were 97.18% and 80.43%, respectively; compared with the original image, the RF classification accuracy was improved by 6.43% and 8.81%, respectively. (2) For this study, the overall classification accuracy of RF in the two regions was positively correlated with the number of random sample points. (3) The image after feature transform was less affected by the number of sample points than the original image. The MNF transform curve of the overall RF classification accuracy in the two regions varied with the number of random sample points but was the smoothest and least affected by the number of sample points, followed by the PCA transform and ICA transform curves. The overall classification accuracies of MNF–RF in the Longkou and Honghu areas did not exceed 0.50% and 3.25%, respectively, with the fluctuation of the number of sample points. This research can provide reference for the fine classification of crops based on UAV-borne hyperspectral images. Full article

The Open Geospatial Consortium (OGC) is an international non-profit standards organization. Established in 1994, OGC aims to make geospatial information and services FAIR-Findable, Accessible, Interoperable, and Reusable. OGC specifications have greatly facilitated interoperability among software, hardware, data, and users in the GIS field. This study collected publications related to OGC specifications from the Web of Science (WoS database) between 1994 to 2020 and conducted a literature analysis using Derwent Data Analyzer and VosViewer, finding that OGC specifications have been widely applied in academic fields. The most productive organizations were Wuhan University and George Mason University; the most common keywords were interoperability, data, and web service. Since 2018, the emerging keywords that have attracted much attention from researchers were 3D city models, 3D modeling, and smart cities. To make geospatial data FAIR, the OGC specifications SWE and WMS served more for “Findable”, SWE contributed more to “Accessible”, WPS and WCS served more for “Interoperable”, and WPS, XML schemas, WFS, and WMS served more for “Reusable”. The OGC specification also serves data and web services for large-scale infrastructure such as the Digital Earth Platform of the Chinese Academy of Sciences. Full article

With the development of tourism and the change in urban functions, the analysis of the spatial pattern of urban tourist flows has become increasingly important. Existing studies have explored and analyzed tourist behavior well, using the appropriate digital footprint data and research methods. However, most studies have ignored internal mechanisms analysis and tourism decision making. This paper proposed a novel framework for tourist behavior analysis inspired by geo-information Tupu, including three modules of the spatiotemporal database, symptom, diagnosis, and implementation. The spatiotemporal database module is mainly used for data acquisition and data cleaning of the digital footprint of tourists. The symptom module is mainly used for revealing the spatial patterns and network structures of tourist flows. The diagnosis and implementation module is mainly used for internal mechanism analysis and tourism decision making under different tourist flow patterns. This paper applied the proposed research framework to Dengfeng City, China, using online travel diaries as the source of digital footprint data, to analyze its tourist behavior. The results were as follows: tourist flows of Dengfeng were unevenly distributed, thus forming an obvious core–periphery structure with intense internal competition and unbalanced power. The difference in tourism resources between its northern and southern areas remains a challenge for future tourism development in Dengfeng. Full article

Commercial and residential spaces are two core types of geographical objects in urban areas. However, these two types of spaces are not independent of each other. Spatial associations exist between them, and a thorough understanding of this spatial association is of great significance for improving the efficiency of urban spatial allocation and realizing scientific spatial planning and governance. Thus, in this paper, the spatial association between commercial and residential spaces in Beijing is quantified with GIS spatial analysis of the average nearest neighbor distance, kernel density, spatial correlation, and honeycomb grid analysis. Point-of-interest (POI) big data of the commercial and residential spaces is used in the quantification since this big data represents a comprehensive sampling of these two spaces. The results show that the spatial distributions of commercial and residential spaces are highly correlated, maintaining a relatively close consumption spatial association. However, the degrees of association between different commercial formats and residential spaces vary, presenting the spatial association characteristics of “integration of daily consumption and separation of nondaily consumption”. The commercial formats of catering services, recreation and leisure services, specialty stores, and agricultural markets are strongly associated with the residential spaces. However, the development of frequently used commercial formats of daily consumption such as living services, convenience stores, and supermarkets appears to lag behind the development of residential spaces. In addition, large-scale comprehensive and specialized commercial formats such as shopping malls, home appliances and electronics stores, and home building materials markets are lagging behind the residential spaces over a wide range. This paper is expected to provide development suggestions for the transformation of urban commercial and residential spaces and the construction of “people-oriented” smart cities. Full article

This paper describes a method for metric recording based on spherical images, which are rectified to document planar surfaces. The proposed method is a multistep workflow in which multiple rectilinear images are (i) extracted from a single spherical projection and (ii) used to recover metric properties. The workflow is suitable for documenting buildings with small and narrow rooms, i.e., documentation projects where the acquisition of 360 images is faster than the traditional acquisition of several photographs. Two different rectification procedures were integrated into the current implementation: (i) an analytical method based on control points and (ii) a geometric procedure based on two sets of parallel lines. Constraints based on line parallelism can be coupled with the focal length of the rectified image to estimate the rectifying transformation. The calculation of the focal length does not require specific calibrations projects. It can be derived from the spherical image used during the documentation project, obtaining a rectified image with just an overall scale ambiguity. Examples and accuracy evaluation are illustrated and discussed to show the pros and cons of the proposed method. Full article

This paper proposes an efficient approach for the plane segmentation of indoor and corridor scenes. Specifically, the proposed method first uses voxels to pre-segment the scene and establishes the topological relationship between neighboring voxels. The voxel normal vectors are projected onto the surface of a Gaussian sphere based on the corresponding directions to achieve fast plane grouping using a variant of the K-means approach. To improve the segmentation integration, we propose releasing the points from the specified voxels and establishing second-order relationships between different primitives. We then introduce a global energy-optimization strategy that considers the unity and pairwise potentials while including high-order sequences to improve the over-segmentation problem. Three benchmark methods are introduced to evaluate the properties of the proposed approach by using the ISPRS benchmark datasets and self-collected in-house. The results of our experiments and the comparisons indicate that the proposed method can return reliable segmentation with precision over 72% even with the low-cost sensor, and provide the best performances in terms of the precision and recall rate compared to the benchmark methods. Full article

Urban change detection is an important part of sustainable urban planning, regional development, and socio-economic analysis, especially in regions with limited access to economic and demographic statistical data. The goal of this research is to create a strategy that enables the extraction of indicators from large-scale orthoimages of different resolution with practically acceptable accuracy after a short training process. Remote sensing data can be used to detect changes in number of buildings, forest areas, and other landscape objects. In this paper, aerial images of a digital raster orthophoto map at scale 1:10,000 of the Republic of Lithuania (ORT10LT) of three periods (2009–2010, 2012–2013, 2015–2017) were analyzed. Because of the developing technologies, the quality of the images differs significantly and should be taken into account while preparing the dataset for training the semantic segmentation model DeepLabv3 with a ResNet50 backbone. In the data preparation step, normalization techniques were used to ensure stability of image quality and contrast. Focal loss for the training metric was selected to deal with the misbalanced dataset. The suggested model training process is based on the transfer learning technique and combines using a model with weights pretrained in ImageNet with learning on coarse and fine-tuning datasets. The coarse dataset consists of images with classes generated automatically from Open Street Map (OSM) data and the fine-tuning dataset was created by manually reviewing the images to ensure that the objects in images match the labels. To highlight the benefits of transfer learning, six different models were trained by combining different steps of the suggested model training process. It is demonstrated that using pretrained weights results in improved performance of the model and the best performance was demonstrated by the model which includes all three steps of the training process (pretrained weights, training on coarse and fine-tuning datasets). Finally, the results obtained with the created machine learning model enable the implementation of different approaches to detect, analyze, and interpret urban changes for policymakers and investors on different levels on a local map, grid, or municipality level. Full article

With the rise of social media platforms, tourists tend to share their experiences in the form of texts, photos, and videos on social media. These user-generated contents (UGC) play an important role in shaping tourism destination images (TDI) and directly affect the decision-making process of tourists. Among UGCs, photos represent tourists’ visual preferences for a specific area. Paying attention to the value of photos, several studies have attempted to analyze them using deep learning technology. However, the research methods that analyze tourism photos using recent deep learning technology have a limitation in that they cannot properly classify unique photos appearing in specific tourist attractions with predetermined photo categories such as Places365 or ImageNet dataset or it takes a lot of time and effort to build a separate training dataset to train the model and to generate a tourism photo classification category according to a specific tourist destination. The purpose of this study is to propose a method of automatically classifying tourist photos by tourist attractions by applying the methods of the image feature vector clustering and the deep learning model. To this end, first, we collected photos attached to reviews posted by foreign tourists on TripAdvisor. Second, we embedded individual images as 512-dimensional feature vectors using the VGG16 network pre-trained with Places365 and reduced them to two dimensions with t-SNE(t-Distributed Stochastic Neighbor Embedding). Then, clusters were extracted through HDBSCAN(Hierarchical Clustering and Density-Based Spatial Clustering of Applications with Noise) analysis and set as a regional image category. Finally, the Siamese Network was applied to remove noise photos within the cluster and classify photos according to the category. In addition, this study attempts to confirm the validity of the proposed method by applying it to two representative tourist attractions such as ‘Gyeongbokgung Palace’ and ‘Insadong’ in Seoul. As a result, it was possible to identify which visual elements of tourist attractions are attractive to tourists. This method has the advantages in that it is not necessary to create a classification category in advance, it is possible to flexibly extract categories for each tourist destination, and it is able to improve classification performance even with a rather small volume of a dataset. Full article

The widespread of shipborne Automatic Identification System (AIS) equipment will continue to produce a large amount of spatiotemporal trajectory data. In order to explore and understand the hidden behaviour patterns in the data, an interactive visual analysis method combining multiple views is proposed. The method mainly includes four parts: using a trajectory compression algorithm that takes into account the vessel motion characteristics to preprocess the vessel trajectory data; displaying and replaying vessel trajectories based on Electronic Chart System (ECS), and proposing a detection algorithm for vessel stay points based on the principle of spatiotemporal density to semantically label vessel trajectories; using the Fast Dynamic Time Warping (FastDTW) similarity measurement algorithm and the Ordering Points to Identify the Clustering Structure (OPTICS) clustering algorithm to cluster vessel trajectories to show the differences and similarities between vessel traffic flows; and showing the distribution of vessels and the variation trend of vessel density based on the vessel heatmap. Based on the AIS data of the Yangtze River Estuary, three cases are used to prove the usefulness and effectiveness of the system in vessel behaviour analysis. Full article

With the improvement of urban infrastructure and the increase in the coverage of high-rise buildings, the demand for location information services inside buildings is becoming more and more urgent. Moreover, indoor path planning, as a prerequisite and basis for realizing path guidance inside buildings, has become a research focus in the field of location services. This makes the accurate planning of indoor paths an urgent problem to be solved at present. This requires dynamic and precise planning from static fuzzy planning, and the corresponding scene converted from a two-dimensional plane to a three-dimensional one. However, most of the existing indoor path planning methods focus on the use of two-dimensional floor plans in buildings to build indoor maps and rely on traditional path search algorithms for pathfinding, which lack in the efficient use of the building’s own geometric and attribute information and lack consideration of the internal spatial topology of the building, making it difficult to meet the needs of indoor multi-layer continuous space path planning. Considering this relationship, it is difficult to meet the path planning needs of indoor multi-layer continuous spaces. In addition, the two-dimensional expression dominated by arrows and line drawings also greatly reduces the intuitiveness and interactivity of path expression. Regarding this, this paper combines the GeoSOT grid with accurate real geographic information and the BIM model and proposes an accurate indoor path planning method. Finally, using Guanlan Commercial Street in Baiyin City as the experimental object, the precise planning and generation of indoor paths and the interaction of visual displays on the web page are realized. It has been verified that the method has certain reference and application values for meeting the demand of location information services in buildings and building an integrated indoor–outdoor navigation service platform. Full article

Applications of machine-learning-based approaches in the geosciences have witnessed a substantial increase over the past few years. Here we present an approach that accounts for spatial autocorrelation by introducing spatial features to the models. In particular, we explore two types of spatial features, namely spatial lag and eigenvector spatial filtering (ESF). These features are used within the widely used random forest (RF) method, and their effect is illustrated on two public datasets of varying sizes (Meuse and California housing datasets). The least absolute shrinkage and selection operator (LASSO) is used to determine the best subset of spatial features, and nested cross-validation is used for hyper-parameter tuning and performance evaluation. We utilize Moran’s I and local indicators of spatial association (LISA) to assess how spatial autocorrelation is captured at both global and local scales. Our results show that RF models combined with either spatial lag or ESF features yield lower errors (up to 33% different) and reduce the global spatial autocorrelation of the residuals (up to 95% decrease in Moran’s I) compared to the RF model with no spatial features. The local autocorrelation patterns of the residuals are weakened as well. Compared to benchmark geographically weighted regression (GWR) models, the RF models with spatial features yielded more accurate models with similar levels of global and local autocorrelation in the prediction residuals. This study reveals the effectiveness of spatial features in capturing spatial autocorrelation and provides a generic machine-learning modelling workflow for spatial prediction. Full article

The quality of street space has attracted attention. It is important to understand the needs of different population groups for street space quality, especially the rapidly growing elderly group. Improving the quality of street space is conducive to promoting the physical leisure activities of the elderly to benefit to their health. Therefore, it is important to evaluate street space quality for the elderly. The existing studies, on the one hand, are limited by the sample size of traditional survey data, which is hard to apply on a large scale; on the other hand, there is a lack of consideration for factors that reveal the quality of street space from the perspective of the elderly. This paper takes Guangzhou as an example to evaluate the quality of street space. First, the sample street images were scored by the elderly on a small scale; then the regression analysis was used to extract the street elements that the elderly care about. Last, the street elements were put into the random forest model to assess street space quality io a large scale. It was found that the green view rate and sidewalks are positively correlated with satisfaction, and the positive effect increases in that order. Roads, buildings, sky, vehicles, walls, ceilings, glass windows, runways, railings, and rocks are negatively correlated with satisfaction, and the negative effect increases in that order. The mean satisfaction score of the quality of street space for the elderly’s recreational physical activities in three central districts of Guangzhou (Yuexiu, Liwan, and Haizhu) is 2.6, among which Xingang street gets the highest quality score (2.92), and Hailong street has the lowest quality score (2.32). These findings are useful for providing suggestions to governors and city designers for street space optimization. Full article

Wind can influence people’s behavior and their way of inhabiting an architectural or urban space. Furthermore, virtual reality (VR) enables the simulation of different physical and sensitive phenomena such as the wind. We aim to analyze the effects of different wind representations in terms of perception of its properties and sense of presence in VR. We carry out two within-subject studies aiming at evaluating different wind representation suggestions (including audiovisual and tactile stimuli) to identify their effects on wind properties’ perception and sense of presence in the VR scene. Our analysis showed significant effects of tactile restitution over the visual effects used in the study, both for understanding wind properties and for increasing the sense of presence in the VR scene. The tactile condition (T) reduced the estimation error of wind direction by 27% compared to the visual condition (V). The wind force error was reduced by 9.8% using (T) with (V). (T) increased the sense of presence by 12.2% compared to (V). Our second experiment showed an overestimation of the wind force perceived compared to the reference value of the Beaufort scale. For the maximum force value evaluated, the average result was 91% higher than the reference value, while for the lower, the average answer was 77% higher than the reference value. Previous studies have evaluated wind rendering in virtual reality, and others have studied the visualization of wind simulation results. To our knowledge, our study is the first to compare the perception of these two types of representations as well as the effects of wind on elements of the context. We also compared the wind perception to a reference-based method, the Beaufort scale. Full article