Spatial Humanities and Geo-computation for Social Sciences:Advances and Applications

Humanities and Social Sciences(HSS) are undergoing the transformation of spatialization and quantification. Geo-computation, with geoinformatics(including RS: Remote Sensing;GIS: Geographical Information System;GNSS: Global Navigation Satellite System), provides effective computational and spatialization methods and tools for HSS. Spatial Humanities and Geo-computation for Social Sciences(SH&GSS) is a field coupling geo-computation, and geoinformatics, with HSS. This special issue accepted a set of contributions highlighting recent advances in methodologies and applications of SH&GSS, which are related to sentiment spatial analysis from social media data, emotional change spatial analysis from news data, spatial analysis of social media related to COVID-19, crime spatiotemporal analysis, “double evaluation” for Land Use/Land Cover(LUCC), Specially Protected Natural Areas(SPNA) analysis, editing behavior analysis of Volunteered Geographic Information(VGI), electricity consumption anomaly detection, First and Last Mile Problem(FLMP) of public transport, and spatial interaction network analysis for crude oil trade network. Based on these related researches, we aim to present an overview of SH&GSS, and propose some future research directions for SH&HSS.

GIScience and remote sensing in natural resource and environmental research:Status quo and future perspectives

Geographic information science(GIScience)and remote sensing have long provided essential data and method-ological support for natural resource challenges and environmental problems research.With increasing advances in information technology,natural resource and environmental science research faces the dual challenges of data and computational intensiveness.Therefore,the role of remote sensing and GIScience in the fields of natural resources and environmental science in this new information era is a key concern of researchers.This study clarifies the definition and frameworks of these two disciplines and discusses their role in natural resource and environmental research.GIScience is the discipline that studies the abstract and formal expressions of the basic concepts and laws of geography,and its research framework mainly consists of geo-modeling,geo-analysis,and geo-computation.Remote sensing is a comprehensive technology that deals with the mechanisms of human ef-fects on the natural ecological environment system by observing the earth surface system.Its main areas include sensors and platforms,information processing and interpretation,and natural resource and environmental appli-cations.GIScience and remote sensing provide data and methodological support for resource and environmental science research.They play essential roles in promoting the development of resource and environmental science and other related technologies.This paper provides forecasts of ten future directions for GIScience and eight future directions for remote sensing,which aim to solve issues related to natural resources and the environment.

Photogrammetry and Deep Learning

Deep learning has become popular and the mainstream technology in many researches related to learning,and has shown its impact on photogrammetry.According to the definition of photogrammetry,that is,a subject that researches shapes,locations,sizes,characteristics and inter-relationships of real objects from optical images,photogrammetry considers two aspects,geometry and semantics.From the two aspects,we review the history of deep learning and discuss its current applications on photogrammetry,and forecast the future development of photogrammetry.In geometry,the deep convolutional neural network(CNN)has been widely applied in stereo matching,SLAM and 3D reconstruction,and has made some effects but needs more improvement.In semantics,conventional methods that have to design empirical and handcrafted features have failed to extract the semantic information accurately and failed to produce types of“semantic thematic map”as 4D productions(DEM,DOM,DLG,DRG)of photogrammetry.This causes the semantic part of photogrammetry be ignored for a long time.The powerful generalization capacity,ability to fit any functions and stability under types of situations of deep leaning is making the automatic production of thematic maps possible.We review the achievements that have been obtained in road network extraction,building detection and crop classification,etc.,and forecast that producing high-accuracy semantic thematic maps directly from optical images will become reality and these maps will become a type of standard products of photogrammetry.At last,we introduce our two current researches related to geometry and semantics respectively.One is stereo matching of aerial images based on deep learning and transfer learning;the other is precise crop classification from satellite spatio-temporal images based on 3D CNN.

Preface to the Album “Digital Photogrammetry and Machine Vision”

Since its birth, photogrammetry technology has made important contributions to topographic mapping, cadastral surveys, and industrial surveys in countries around the world. In the 1970s, the development of digital photography and computer technology represented by digital imaging devices promoted the revolution of photogrammetry theory and technology, and formed the current digital photogrammetry theory and system. In recent years, the rapid development of machine vision and artificial intelligence has had a profound impact and challenge on digital photogrammetry.

LuoJiaAI:A cloud-based artificial intelligence platform for remote sensing image interpretation

The rapid processing,analysis,and mining of remote-sensing big data based on intelligent interpretation technology using remote-sensing cloud computing platforms(RS-CCPs)have recently become a new trend.The existing RS-CCPs mainly focus on developing and optimizing high-performance data storage and intelligent computing for common visual representation,which ignores remote sensing data characteristics such as large image size,large-scale change,multiple data channels,and geographic knowledge embedding,thus impairing computational efficiency and accuracy.We construct a LuoJiaAI platform composed of a standard large-scale sample database(LuoJiaSET)and a dedicated deep learning framework(LuoJiaNET)to achieve state-of-the-art performance on five typical remote sensing interpretation tasks,including scene classification,object detection,land-use classification,change detection,and multi-view 3D reconstruction.The details of the LuoJiaAI application experiment can be found at the white paper for LuoJiaAI industrial application.In addition,LuoJiaAI is an open-source RS-CCP that supports the latest Open Geospatial Consortium(OGC)standards for better developing and sharing Earth Artificial Intelligence(AI)algorithms and products on benchmark datasets.LuoJiaAI narrows the gap between the sample database and deep learning frameworks through a user-friendly data-framework collaboration mechanism,showing great potential in high-precision remote sensing mapping applications.

Luojia-HSSR:A high spatial-spectral resolution remote sensing dataset for land-cover classification with a new 3D-HRNet

High Spatial and Spectral Resolution(HSSR)remote-sensing images can provide rich spectral bands and detailed ground information,but there is a relative lack of research on this new type of remote-sensing data.Although there are already some HSSR datasets for deep learning model training and testing,the data volume of these datasets is small,resulting in low classification accuracy and weak generalization ability of the trained models.In this paper,an HSSR dataset Luojia-HSSR is constructed based on aerial hyperspectral imagery of southern Shenyang City of Liaoning Province in China.To our knowledge,it is the largest HSSR dataset to date,with 6438 pairs of 256×256 sized samples(including 3480 pairs in the training set,2209 pairs in the test set,and 749 pairs in the validation set),covering area of 161 km2 with spatial resolution 0.75 m,249 Visible and Near-Infrared(VNIR)spectral bands,and corresponding to 23 classes of field-validated ground coverage.It is an ideal experimental data for spatial-spectral feature extraction.Furthermore,a new deep learning model 3D-HRNet for interpreting HSSR images is proposed.The conv-neck in HRNet is modified to better mine the spatial information of the images.Then,a 3D convolution module with attention mechanism is designed to capture the global-local fine spectral information simultaneously.Subsequently,the 3D convolution is inserted into the HRNet to optimize the performance.The experiments show that the 3D-HRNet model has good interpreting ability for the Luojia-HSSR dataset with the Frequency Weighted Intersection over Union(FWIoU)reaching 80.54%,indicating that the Luojia-HSSR dataset constructed in this paper and the proposed 3D-HRnet model have good applicable prospects for processing HSSR remote sensing images.

Advances in urban information extraction from high-resolution remote sensing imagery

The study of urban area is one of the hottest research topics in the field of remote sensing. With the accumulation of high-resolution(HR) remote sensing data and emerging of new satellite sensors, HR observation of urban areas has become increasingly possible, which provides us with more elaborate urban information. However, the strong heterogeneity in the spectral and spatial domain of HR imagery brings great challenges to urban remote sensing. In recent years, numerous approaches were proposed to deal with HR image interpretation over complex urban scenes, including a series of features from low level to high level, as well as state-of-the-art methods depicting not only the urban extent, but also the intra-urban variations. In this paper, we aim to summarize the major advances in HR urban remote sensing from the aspects of feature representation and information extraction. Moreover, the future trends are discussed from the perspectives of methodology, urban structure and pattern characterization, big data challenge, and global mapping.

High-resolution urban land-cover mapping and landscape analysis of the 42 major cities in China using ZY-3 satellite images

Detailed and precise urban land-cover maps are crucial for urban-related studies. However, there are limited ways of mapping high-resolution urban land cover over large areas. In this paper, we propose an operational framework to map urban land cover on the basis of Ziyuan-3 satellite images. Based on this framework, we produced the first high-resolution(2 m) urban land-cover map(Hi-ULCM) covering the 42 major cities of China. The overall accuracy of the Hi-ULCM dataset is 88.55%, of which 14 cities have an overall accuracy of over 90%. Most of the producer’s accuracies and user’s accuracies of the land-cover classes exceed 85%. We further conducted a landscape pattern analysis in the 42 cities based on Hi-ULCM. In terms of the comparison between the 42 cities in China, we found that the difference in the land-cover composition of urban areas is related to the climatic characteristics and urbanization levels, e.g., cities with warm climates generally have higher proportions of green spaces. It is also interesting to find that cities with higher urbanization levels are more habitable, in general. From the landscape viewpoint, the geometric complexity of the landscape increases with the urbanization level.Compared with the existing medium-resolution land-cover/use datasets(at a 30-m resolution), HiULCM represents a significant advance in accurately depicting the detailed land-cover footprint within the urban areas of China, and will be of great use for studies of urban ecosystems.

Development,application,and prospects for Chinese land observation satellites

The launching of CBERS-01(China Brazil Earth Resource Satellite)in 1999,China’s first land observation satellite,signifies an unprecedented milestone in Chinese satellite remote sensing history.Since then,a large number of applications have been developed that drew upon solely CBERS-01 and other Chinese land observation satellites.The application development evolves from one satellite to multiple satellites,from one series of satellites to multiple series,from scientific research to industrial applications.Six aspects of the Chinese land observation satellite program are discussed in this paper:development status,data sharing and distribution,satellite calibration,industrial data applications,future prospects,and conclusion.

A virtual globe-based vector data model:quaternary quadrangle vector tile model

This study proposes a virtual globe-based vector data model named the quaternary quadrangle vector tile model(QQVTM)in order to better manage,visualize,and analyze massive amounts of global multi-scale vector data.The model integrates the quaternary quadrangle mesh(a discrete global grid system)and global image,terrain,and vector data.A QQVTM-based organization method is presented to organize global multi-scale vector data,including linear and polygonal vector data.In addition,tilebased reconstruction algorithms are designed to search and stitch the vector fragments scattered in tiles to reconstruct and store the entire vector geometries to support vector query and 3D analysis of global datasets.These organized vector data are in turn visualized and queried using a geometry-based approach.Our experimental results demonstrate that the QQVTM can satisfy the requirements for global vector data organization,visualization,and querying.Moreover,the QQVTM performs better than unorganized 2D vectors regarding rendering efficiency and better than the latitude–longitude-based approach regarding data redundancy.

Layout graph model for semantic façade reconstruction using laser point clouds

Building façades can feature different patterns depending on the architectural style,function-ality,and size of the buildings;therefore,reconstructing these façades can be complicated.In particular,when semantic façades are reconstructed from point cloud data,uneven point density and noise make it difficult to accurately determine the façade structure.When inves-tigating façade layouts,Gestalt principles can be applied to cluster visually similar floors and façade elements,allowing for a more intuitive interpretation of façade structures.We propose a novel model for describing façade structures,namely the layout graph model,which involves a compound graph with two structure levels.In the proposed model,similar façade elements such as windows are first grouped into clusters.A down-layout graph is then formed using this cluster as a node and by combining intra-and inter-cluster spacings as the edges.Second,a top-layout graph is formed by clustering similar floors.By extracting relevant parameters from this model,we transform semantic façade reconstruction to an optimization strategy using simulated annealing coupled with Gibbs sampling.Multiple façade point cloud data with different features were selected from three datasets to verify the effectiveness of this method.The experimental results show that the proposed method achieves an average accuracy of 86.35%.Owing to its flexibility,the proposed layout graph model can deal with different types of façades and qualities of point cloud data,enabling a more robust and accurate reconstruc-tion of façade models.

Geoinformatics education in China

The paper gives an overview of the current status of education in geoinformatics in China.First,the paper provides a brief introduction to the history of geoinformatics education in China and a general review of the scientific and technological development.It then presents how the development affects the education and training in China.In the paper,universities and institutes in China that can award academic degrees related to geoinformatics are summarized,and undergraduate majors are briefly introduced.Next,the paper reports the work having been done by the national expert group on Surveying and Mapping,including the revision of discipline catalog and guide for graduate education and requirements.A list of typical curricula in geoinformatics education is suggested.Activities on promoting the graduate student exchange platform are presented.Finally,a case study of geoinformatics education in Wuhan University is discussed.

A scalable cyberinfrastructure and cloud computing platform for forest aboveground biomass estimation based on the Google Earth Engine

Earth observation(EO)data,such as high-resolution satellite imagery or LiDAR,has become one primary source for forests Aboveground Biomass(AGB)mapping and estimation.However,managing and analyzing the large amount of globally or locally available EO data remains a great challenge.The Google Earth Engine(GEE),which leverages cloud-computing services to provide powerful capabilities on the management and rapid analysis of various types of EO data,has appeared as an inestimable tool to address this challenge.In this paper,we present a scalable cyberinfrastructure for on-the-fly AGB estimation,statistics,and visualization over a large spatial extent.This cyberinfrastructure integrates state-of-the-art cloud computing applications,including GEE,Fusion Tables,and the Google Cloud Platform(GCP),to establish a scalable,highly extendable,and highperformance analysis environment.Two experiments were designed to demonstrate its superiority in performance over the traditional desktop environment and its scalability in processing complex workflows.In addition,a web portal was developed to integrate the cyberinfrastructure with some visualization tools(e.g.Google Maps,Highcharts)to provide a Graphical User Interfaces(GUI)and online visualization for both general public and geospatial researchers.

Geoprocessing in Cloud Computing platforms-a comparative analysis

The emergence of Cloud Computing technologies brings a new information infrastructure to users.Providing geoprocessing functions in Cloud Computing platforms can bring scalable,on-demand,and cost
effective geoprocessing services to geospatial users.This paper provides a comparative analysis of geoprocessing in Cloud Computing platforms
Microsoft Windows Azure and Google App Engine.The analysis compares differences in the data storage,architecture model,and development environment based on the experience to develop geoprocessing services in the two Cloud Computing platforms;emphasizes the importance of virtualization;recommends applications of hybrid geoprocessing Clouds,and suggests an interoperable solution on geoprocessing Cloud services.The comparison allows one to selectively utilize Cloud Computing platforms or hybrid Cloud pattern,once it is understood that the current development of geoprocessing Cloud services is restricted to specific Cloud Computing platforms with certain kinds of technologies.The performance evaluation is also performed over geoprocessing services deployed in public Cloud platforms.The tested services are developed using geoprocessing algorithms from different vendors,GeoSurf and Java Topology Suite.The evaluation results provide a valuable reference on providing elastic and cost-effective geoprocessing Cloud services.

Geoinformatics education and outreach:looking forward

Geoinformatics education is a key factor for sustainable development of geo-spatial sciences and industries.There have been a variety of educational activities focusing on education and training,technology transfer,and capability building in photogrammetry,remote sensing,and spatial information science,together known as Geoinformatics.Geoinformatics education is an essential mission and even determinant in the ISPRS society.The paper discusses key issues in Geoinformatics education.It reviews educational activities from the ISPRS perspective and summarizes lessons learned from these actions.A vision towards future trends of Geoinformatics education in the ISPRS is provided.

Adaptive polarimetric decomposition using incoherent ground scattering models without reflection symmetry assumption

Most of the current incoherent polarimetric decompositions employ coherent models to describe ground scattering;however,this cannot truly reflect the fact especially in natural ground surfaces.This paper proposes a highly adaptive decomposition with incoherent ground scattering models(ADIGSM).In ADIGSM,Neumann’s adaptive model is employed to describe volume scattering,and to explain cross-polarized power in remainder matrix,so that we can obtain orientation angle randomness for both volume scattering and the dominant ground scattering.The computation of volume scattering parameters is strictly constrained for non-negative eigenvalues,while the volume scattering parameters that explain the most cross-polarized power are selected.When applying ADIGSM to NASA’s UAVSAR data,the negative component powers were obtained in quite a few forest pixels.Compared with several newest decompositions,the volume scattering power is obviously lowered,especially in areas dominated by surface scattering or double bounce scattering.The orientation angle randomness of each component is reasonable as well.ADIGSM has potential to be applied in the fields such as PolSAR image classification,land cover mapping,speckle filtering,soil moisture and roughness estimation,etc.

Indoor scene texturing based on single mobile phone images and 3D model fusion

Realistic texture mapping and coherent up-to-date rendering is one of the most important issues in indoor 3-D modelling.However,existing texturing approaches are usually performed manually during the modelling process,and cannot accommodate changes in indoor environments occurring after the model was created,resulting in outdated and misleading texture rendering.In this study,a structured learning-based texture mapping method is proposed for automatic mapping a single still photo from a mobile phone onto an alreadyconstructed indoor 3-D model.The up-to-date texture is captured using a smart phone,and the indoor structural layout is extracted by incorporating per-pixel segmentation in the FCN algorithm and the line constraints into a structured learning algorithm.This enables real-time texture mapping according to parts of the model,based on the structural layout.Furthermore,the rough camera pose is estimated by pedestrian dead reckoning(PDR)and map information to determine where to map the texture.The experimental results presented in this paper demonstrate that our approach can achieve accurate fusion of 3-D triangular meshes with 2-D single images,achieving low-cost and automatic indoor texture updating.Based on this fusion approach,users can have a better experience in virtual indoor3-D applications.