Special Issue "Machine Learning for Geospatial Data Analysis"

A special issue of ISPRS International Journal of Geo-Information (ISSN 2220-9964).

Deadline for manuscript submissions: closed (28 February 2018)

Special Issue Editors

Guest Editor
Dr. Jan Dirk Wegner

Department of Civil, Environmental and Geomatic Engineering, Institute of Geodesy and Photogrammetry, ETH Zürich, Stefano-Franscini-Platz 5, 8093 Zürich, Switzerland
Website | E-Mail
Interests: geospatial computer vision; photogrammetry; remote sensing; large-scale machine learning; deep convolutional neural networks
Guest Editor
Prof. Ribana Roscher

University of Bonn, Photogrammetry, IGG, Nussallee 15, 53115 Bonn, Germany
Website | E-Mail
Interests: landcover classification methods; sparse representation for large scale remote sensing data; incremental/sequential learning; feature/representation learning
Guest Editor
Dr. Michele Volpi

Swiss Data Science Center, ETH Zurich, Switzerland
Website | E-Mail
Interests: very high resolution remote sensing; multitemporal image processing; object detection in remote sensing data; machine learning; deep learning; geosciences
Guest Editor
Dr. Fabio Veronesi

Harper Adams University
Website | E-Mail
Interests: spatial data analysis; spatial statistics; geostatistics; machine learning; GIS; environmental statistics

Special Issue Information

Dear Colleagues,

With this Special Issue on "Machine Learning for Geospatial Data Analysis" we aim at fostering collaboration between the Remote Sensing, GIScience, Computer Vision, and Machine Learning communities.

The interpretation of Big GeoData calls for highly automated approaches relying on new machine learning and data mining approaches. Extraction of meaningful information at large scale from heterogeneous, georeferenced data is a major research topic in quantitative geography, remote sensing, GIScience, cartography, geospatial computer vision, and machine learning. We invite original works to this Special Issue that apply machine learning to such diverse data, ranging from georeferenced imagery and point clouds, georeferenced text corpora and text sources, GIS databases, large-scale (online) maps or any combination of these. Data can either be acquired with dedicated (imaging) campaigns or be collected from crowd-sourced, publicly available data sets like openstreetmap or Mapillary. A major aspect is the joint processing of such data for information extraction. Topics include, but are not limited to:

  • Object reconstruction, recognition, and classification at large scale
  • Supervised, weakly supervised, transfer, and human-in-the-loop learning
  • Joint GIS and image interpretation
  • Big GeoData mining
  • Applications to cities, autonomous driving, rapid hazard response, vegetation and landscape mapping.

Prospective authors are cordially invited to contribute to this theme issue by submitting an original article that deals with one of the sub-fields until 31 January 2018. All submitting authors are strongly encouraged to test their method on a relevant benchmark data set, to compare against baseline approaches and to publicly release source code and potentially the data used in the paper, on acceptance.

Dr. Jan Dirk Wegner
Prof. Ribana Roscher
Dr. Michele Volpi
Dr. Fabio Veronesi
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All papers will be peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. ISPRS International Journal of Geo-Information is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1000 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • Big GeoData mining
  • Object reconstruction, recognition, and classification at large scale
  • Supervised, weakly supervised, transfer, and human-in-the-loop learning
  • Joint GIS, geo-located image and text interpretation
  • Applications to cities, autonomous driving, rapid hazard response, vegetation mapping, natural and human-induced phenomenon monitoring.

Published Papers (7 papers)

View options order results:
result details:
Displaying articles 1-7
Export citation of selected articles as:

Editorial

Jump to: Research

Open AccessEditorial Foreword to the Special Issue on Machine Learning for Geospatial Data Analysis
ISPRS Int. J. Geo-Inf. 2018, 7(4), 147; https://doi.org/10.3390/ijgi7040147
Received: 9 April 2018 / Revised: 9 April 2018 / Accepted: 10 April 2018 / Published: 13 April 2018
PDF Full-text (186 KB) | HTML Full-text | XML Full-text
Abstract
Advances in machine learning research are pushing the limits of geographical information sciences (GIScience) by offering accurate procedures to analyze small-to-big GeoData. This Special Issue groups together six original contributions in the field of GeoData-driven GIScience that focus mainly on three different areas:
[...] Read more.
Advances in machine learning research are pushing the limits of geographical information sciences (GIScience) by offering accurate procedures to analyze small-to-big GeoData. This Special Issue groups together six original contributions in the field of GeoData-driven GIScience that focus mainly on three different areas: extraction of semantic information from satellite imagery, image recommendation, and map generalization. Different technical approaches are chosen for each sub-topic, from deep learning to latent topic models. Full article
(This article belongs to the Special Issue Machine Learning for Geospatial Data Analysis)

Research

Jump to: Editorial

Open AccessFeature PaperArticle Multi-Temporal Land Cover Classification with Sequential Recurrent Encoders
ISPRS Int. J. Geo-Inf. 2018, 7(4), 129; https://doi.org/10.3390/ijgi7040129
Received: 22 January 2018 / Revised: 13 March 2018 / Accepted: 17 March 2018 / Published: 21 March 2018
Cited by 1 | PDF Full-text (7805 KB) | HTML Full-text | XML Full-text
Abstract
Earth observation (EO) sensors deliver data at daily or weekly intervals. Most land use and land cover classification (LULC) approaches, however, are designed for cloud-free and mono-temporal observations. The increasing temporal capabilities of today’s sensors enable the use of temporal, along with spectral
[...] Read more.
Earth observation (EO) sensors deliver data at daily or weekly intervals. Most land use and land cover classification (LULC) approaches, however, are designed for cloud-free and mono-temporal observations. The increasing temporal capabilities of today’s sensors enable the use of temporal, along with spectral and spatial features.Domains such as speech recognition or neural machine translation, work with inherently temporal data and, today, achieve impressive results by using sequential encoder-decoder structures. Inspired by these sequence-to-sequence models, we adapt an encoder structure with convolutional recurrent layers in order to approximate a phenological model for vegetation classes based on a temporal sequence of Sentinel 2 (S2) images. In our experiments, we visualize internal activations over a sequence of cloudy and non-cloudy images and find several recurrent cells that reduce the input activity for cloudy observations. Hence, we assume that our network has learned cloud-filtering schemes solely from input data, which could alleviate the need for tedious cloud-filtering as a preprocessing step for many EO approaches. Moreover, using unfiltered temporal series of top-of-atmosphere (TOA) reflectance data, our experiments achieved state-of-the-art classification accuracies on a large number of crop classes with minimal preprocessing, compared to other classification approaches. Full article
(This article belongs to the Special Issue Machine Learning for Geospatial Data Analysis)
Figures

Graphical abstract

Open AccessArticle Generative Street Addresses from Satellite Imagery
ISPRS Int. J. Geo-Inf. 2018, 7(3), 84; https://doi.org/10.3390/ijgi7030084
Received: 9 January 2018 / Revised: 13 February 2018 / Accepted: 17 February 2018 / Published: 8 March 2018
Cited by 1 | PDF Full-text (89783 KB) | HTML Full-text | XML Full-text
Abstract
We describe our automatic generative algorithm to create street addresses from satellite images by learning and labeling roads, regions, and address cells. Currently, 75% of the world’s roads lack adequate street addressing systems. Recent geocoding initiatives tend to convert pure latitude and longitude
[...] Read more.
We describe our automatic generative algorithm to create street addresses from satellite images by learning and labeling roads, regions, and address cells. Currently, 75% of the world’s roads lack adequate street addressing systems. Recent geocoding initiatives tend to convert pure latitude and longitude information into a memorable form for unknown areas. However, settlements are identified by streets, and such addressing schemes are not coherent with the road topology. Instead, we propose a generative address design that maps the globe in accordance with streets. Our algorithm starts with extracting roads from satellite imagery by utilizing deep learning. Then, it uniquely labels the regions, roads, and structures using some graph- and proximity-based algorithms. We also extend our addressing scheme to (i) cover inaccessible areas following similar design principles; (ii) be inclusive and flexible for changes on the ground; and (iii) lead as a pioneer for a unified street-based global geodatabase. We present our results on an example of a developed city and multiple undeveloped cities. We also compare productivity on the basis of current ad hoc and new complete addresses. We conclude by contrasting our generative addresses to current industrial and open solutions. Full article
(This article belongs to the Special Issue Machine Learning for Geospatial Data Analysis)
Figures

Figure 1

Open AccessFeature PaperArticle Classification of PolSAR Images by Stacked Random Forests
ISPRS Int. J. Geo-Inf. 2018, 7(2), 74; https://doi.org/10.3390/ijgi7020074
Received: 30 January 2018 / Revised: 16 February 2018 / Accepted: 18 February 2018 / Published: 23 February 2018
Cited by 1 | PDF Full-text (44702 KB) | HTML Full-text | XML Full-text
Abstract
This paper proposes the use of Stacked Random Forests (SRF) for the classification of Polarimetric Synthetic Aperture Radar images. SRF apply several Random Forest instances in a sequence where each individual uses the class estimate of its predecessor as an additional feature. To
[...] Read more.
This paper proposes the use of Stacked Random Forests (SRF) for the classification of Polarimetric Synthetic Aperture Radar images. SRF apply several Random Forest instances in a sequence where each individual uses the class estimate of its predecessor as an additional feature. To this aim, the internal node tests are designed to work not only directly on the complex-valued image data, but also on spatially varying probability distributions and thus allow a seamless integration of RFs within the stacking framework. Experimental results show that the classification performance is consistently improved by the proposed approach, i.e., the achieved accuracy is increased by 4 % and 7 % for one fully- and one dual-polarimetric dataset. This increase only comes at the cost of a linear increased training and prediction time, which is rather limited as the method converges quickly. Full article
(This article belongs to the Special Issue Machine Learning for Geospatial Data Analysis)
Figures

Graphical abstract

Open AccessArticle A Space-Time Periodic Task Model for Recommendation of Remote Sensing Images
ISPRS Int. J. Geo-Inf. 2018, 7(2), 40; https://doi.org/10.3390/ijgi7020040
Received: 31 October 2017 / Revised: 14 January 2018 / Accepted: 21 January 2018 / Published: 29 January 2018
Cited by 1 | PDF Full-text (4741 KB) | HTML Full-text | XML Full-text
Abstract
With the rapid development of remote sensing technology, the quantity and variety of remote sensing images are growing so quickly that proactive and personalized access to data has become an inevitable trend. One of the active approaches is remote sensing image recommendation, which
[...] Read more.
With the rapid development of remote sensing technology, the quantity and variety of remote sensing images are growing so quickly that proactive and personalized access to data has become an inevitable trend. One of the active approaches is remote sensing image recommendation, which can offer related image products to users according to their preference. Although multiple studies on remote sensing retrieval and recommendation have been performed, most of these studies model the user profiles only from the perspective of spatial area or image features. In this paper, we propose a spatiotemporal recommendation method for remote sensing data based on the probabilistic latent topic model, which is named the Space-Time Periodic Task model (STPT). User retrieval behaviors of remote sensing images are represented as mixtures of latent tasks, which act as links between users and images. Each task is associated with the joint probability distribution of space, time and image characteristics. Meanwhile, the von Mises distribution is introduced to fit the distribution of tasks over time. Then, we adopt Gibbs sampling to learn the random variables and parameters and present the inference algorithm for our model. Experiments show that the proposed STPT model can improve the capability and efficiency of remote sensing image data services. Full article
(This article belongs to the Special Issue Machine Learning for Geospatial Data Analysis)
Figures

Figure 1

Open AccessArticle Machine Learning Classification of Buildings for Map Generalization
ISPRS Int. J. Geo-Inf. 2017, 6(10), 309; https://doi.org/10.3390/ijgi6100309
Received: 3 August 2017 / Revised: 7 October 2017 / Accepted: 11 October 2017 / Published: 18 October 2017
Cited by 2 | PDF Full-text (6212 KB) | HTML Full-text | XML Full-text
Abstract
A critical problem in mapping data is the frequent updating of large data sets. To solve this problem, the updating of small-scale data based on large-scale data is very effective. Various map generalization techniques, such as simplification, displacement, typification, elimination, and aggregation, must
[...] Read more.
A critical problem in mapping data is the frequent updating of large data sets. To solve this problem, the updating of small-scale data based on large-scale data is very effective. Various map generalization techniques, such as simplification, displacement, typification, elimination, and aggregation, must therefore be applied. In this study, we focused on the elimination and aggregation of the building layer, for which each building in a large scale was classified as “0-eliminated,” “1-retained,” or “2-aggregated.” Machine-learning classification algorithms were then used for classifying the buildings. The data of 1:1000 scale and 1:25,000 scale digital maps obtained from the National Geographic Information Institute were used. We applied to these data various machine-learning classification algorithms, including naive Bayes (NB), decision tree (DT), k-nearest neighbor (k-NN), and support vector machine (SVM). The overall accuracies of each algorithm were satisfactory: DT, 88.96%; k-NN, 88.27%; SVM, 87.57%; and NB, 79.50%. Although elimination is a direct part of the proposed process, generalization operations, such as simplification and aggregation of polygons, must still be performed for buildings classified as retained and aggregated. Thus, these algorithms can be used for building classification and can serve as preparatory steps for building generalization. Full article
(This article belongs to the Special Issue Machine Learning for Geospatial Data Analysis)
Figures

Figure 1

Open AccessArticle Contextual Building Selection Based on a Genetic Algorithm in Map Generalization
ISPRS Int. J. Geo-Inf. 2017, 6(9), 271; https://doi.org/10.3390/ijgi6090271
Received: 6 July 2017 / Revised: 18 August 2017 / Accepted: 28 August 2017 / Published: 30 August 2017
Cited by 1 | PDF Full-text (6221 KB) | HTML Full-text | XML Full-text
Abstract
In map generalization, scale reduction and feature symbolization inevitably generate problems of overlapping objects or map congestion. To solve the legibility problem with respect to the generalization of dispersed rural buildings, selection of buildings is necessary and can be transformed into an optimization
[...] Read more.
In map generalization, scale reduction and feature symbolization inevitably generate problems of overlapping objects or map congestion. To solve the legibility problem with respect to the generalization of dispersed rural buildings, selection of buildings is necessary and can be transformed into an optimization problem. In this paper, an improved genetic algorithm for building selection is designed to be able to incorporate cartographic constraints related to the building selection problem. Part of the local constraints for building selection is used to constrain the encoding and genetic operation. To satisfy other local constraints, a preparation phase is necessary before building selection, which includes building enlargement, local displacement, conflict detection, and attribute enrichment. The contextual constraints are used to ascertain a fitness function. The experimental results indicate that the algorithm proposed in this article can obtain good results for building selection whilst preserving the spatial distribution characteristics of buildings. Full article
(This article belongs to the Special Issue Machine Learning for Geospatial Data Analysis)
Figures

Figure 1

Back to Top