Special Issue "Object Based Image Analysis for Remote Sensing"

A special issue of Remote Sensing (ISSN 2072-4292). This special issue belongs to the section "Remote Sensing Image Processing".

Deadline for manuscript submissions: 3 March 2020.

Special Issue Editors

Dr. Peter Hofmann
E-Mail Website
Guest Editor
Deggendorf Institute of Technology IAI - Institute for Applied Informatics Technology Campus Freyung Grafenauer Str. 22, D-94078 Freyung, Germany
Tel. +49 8551 91764-29
Interests: Image Analysis, Object Based Image Analysis (OBIA), Artificial Intelligence in Remote Sensing and Geoinformatics, UAS, Urban Remote Sensing, Remote Sensing of Environment
Special Issues and Collections in MDPI journals
Dr. Raul Queiroz Feitosa
E-Mail Website
Guest Editor
Department of Electrical Engineering, Pontifical Catholic University of Rio de Janeiro, r. Marquês de São Vicente, 225, 22451-900, Rio de Janeiro, RJ, Brazil
Tel. +55 21 35271212; Fax: +55 21 35271232
Interests: pattern recognition for remote sensing; image analysis; remote sensing applications
Special Issues and Collections in MDPI journals
Prof. Cláudia Maria de Almeida
E-Mail Website
Guest Editor
National Institute for Space Research, General Coordination for Earth Observation—Division for Remote Sensing, Av. dos Astronautas, 1758, SERE/DSR, Room 06, Jardim da Granja, 12227010 - São José dos Campos, SP, Brazil, PO Box: 515
Tel. +55 12 3208-6428; Fax: +55 12 3208-6488
Interests: urban remote sensing, GISciences, object-based image analysis OBIA, town planning, cellular automata, urban modelling, high spatial resolution sensors, airborne laser scanner (ALS), terrestrial laser scanner (TLS)
Prof. Gilson Alexandre Ostwald Pedro da Costa
E-Mail Website
Guest Editor
Rio de Janeiro State University (UERJ), Institute of Mathematics and Statistics (IME), Department of Informatics and Computer Science. R. São Francisco Xavier, 524, Pavilhão Reitor João Lyra Filho, 6º andar, Sala 6019, Bloco B, 20550-900, Rio de Janeiro, RJ, Brasil
Tel. +55 21 23340144; Fax: +55 21 23340144
Interests: remote sensing, object-based image analysis (OBIA), computer vision, machine learning, deep learning

Special Issue Information

Dear Colleagues,

Object-based Image Analysis (OBIA) has evolved to a widespread methodology for image analysis, especially in the context of remote sensing. With the emergence of Very High Resolution (VHR) remote sensing data it turned out that methods which operate on image segments instead of single pixels show lots of advantages when analyzing the content of remote sensing data. With the advent of user friendly software, which allowed to analyze remote sensing data in OBIA manner, OBIA has been further boosted in the remote sensing and GIS community. The numerous scientific publications dealing with OBIA in the remote sensing and GIS domain show, that this methodology has meanwhile established – probably even as a paradigm for image analysis. Simultaneously, the methodology itself underwent a step-by-step evolution, comprising the development of new segmentation methods, the integration of new classification methods and the development of new methods for change detection and monitoring, just to name a few. Meanwhile, the integration of new methods - mainly originating in AI - plays an important role for OBIA. Exemplary, the explicit formulation and management of knowledge, the application of artificial learning and learning mechanisms, but also self-organizing agent-based systems are interesting new developments in OBIA which originate in AI.

In this special issue, we first intend to outline the state-of-the-art in OBIA for remote sensing and the methodologies it comprises meanwhile. Further, we intend to present recent concepts, frameworks and new methods which found their way to OBIA in conjunction with recent applications and success stories of OBIA in remote sensing. This will span a wide spectrum ranging from: image segmentation methods, software engineering in the context of OBIA, semantic modelling and reasoning, ontologies and knowledge representations, classification methods including Complex Neural Networks (CNNs) to self-organizing approaches such as multi-agent systems. Further, OBIA-specific approaches of change detection and monitoring as well as the incorporation of non-remote sensing and even unstructured data are further aspects we want to deal with. Last but not least cloud computing and Big Earth Data in the context of OBIA are challenging fields we would like to spot at.

We would like to invite colleagues to submit articles about their recent research on any of the following topics but not restricted to:

  • Image segmentation and joined aspects, such as optimization, quality assessment, transferability, etc.
  • Software development and engineering in the context of OBIA including robustness and quality assessment.
  • Knowledge representation and management, including ontologies and reasoning.
  • Classification methods including CNNs and other ANN-based methods.
  • Self-organizing methods such as Multi-Agent Systems in OBIA.
  • Object-based change detection and monitoring methods.
  • Data integration and usage.
  • Cloud computing and Big Earth Data in OBIA.
  • Applications of OBIA in remote sensing and success stories with OBIA.

Prof. Dr. Raul Queiroz Feitosa
Dr. Peter Hofmann
Prof. Cláudia Maria de Almeida
Prof. Gilson Alexandre Ostwald Pedro da Costa
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. Remote Sensing is an international peer-reviewed open access semimonthly 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 1800 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

  • OBIA
  • GeOBIA
  • remote sensing
  • image analysis
  • artificial intelligence
  • knowledge representation
  • big earth data

Published Papers (4 papers)

Order results
Result details
Select all
Export citation of selected articles as:

Research

Open AccessArticle
Identifying Vegetation in Arid Regions Using Object-Based Image Analysis with RGB-Only Aerial Imagery
Remote Sens. 2019, 11(19), 2308; https://doi.org/10.3390/rs11192308 - 03 Oct 2019
Abstract
Vegetation state is usually assessed by calculating vegetation indices (VIs) derived from remote sensing systems where the near infrared (NIR) band is used to enhance the vegetation signal. However VIs are pixel-based and require both visible and NIR bands. Yet, most archived photographs [...] Read more.
Vegetation state is usually assessed by calculating vegetation indices (VIs) derived from remote sensing systems where the near infrared (NIR) band is used to enhance the vegetation signal. However VIs are pixel-based and require both visible and NIR bands. Yet, most archived photographs were obtained with cameras that record only the three visible bands. Attempts to construct VIs with the visible bands alone have shown only limited success, especially in drylands. The current study identifies vegetation patches in the hyperarid Israeli desert using only the visible bands from aerial photographs by adapting an alternative geospatial object-based image analysis (GEOBIA) routine, together with recent improvements in preprocessing. The preprocessing step selects a balanced threshold value for image segmentation using unsupervised parameter optimization. Then the images undergo two processes: segmentation and classification. After tallying modeled vegetation patches that overlap true tree locations, both true positive and false positive rates are obtained from the classification and receiver operating characteristic (ROC) curves are plotted. The results show successful identification of vegetation patches in multiple zones from each study area, with area under the ROC curve values between 0.72 and 0.83. Full article
(This article belongs to the Special Issue Object Based Image Analysis for Remote Sensing)
Show Figures

Graphical abstract

Open AccessArticle
Object-Based Flood Analysis Using a Graph-Based Representation
Remote Sens. 2019, 11(16), 1883; https://doi.org/10.3390/rs11161883 - 12 Aug 2019
Abstract
The amount of freely available satellite data is growing rapidly as a result of Earth observation programmes, such as Copernicus, an initiative of the European Space Agency. Analysing these huge amounts of geospatial data and extracting useful information is an ongoing pursuit. This [...] Read more.
The amount of freely available satellite data is growing rapidly as a result of Earth observation programmes, such as Copernicus, an initiative of the European Space Agency. Analysing these huge amounts of geospatial data and extracting useful information is an ongoing pursuit. This paper presents an alternative method for flood detection based on the description of spatio-temporal dynamics in satellite image time series (SITS). Since synthetic aperture radar (SAR) satellite data has the capability of capturing images day and night, irrespective of weather conditions, it is the preferred tool for flood mapping from space. An object-based approach can limit the necessary computer power and computation time, while a graph-based approach allows for a comprehensible interpretation of dynamics. This method proves to be a useful tool to gain insight in a flood event. Graph representation helps to identify and locate entities within the study site and describe their evolution throughout the time series. Full article
(This article belongs to the Special Issue Object Based Image Analysis for Remote Sensing)
Show Figures

Graphical abstract

Open AccessArticle
Object-Based Window Strategy in Thermal Sharpening
Remote Sens. 2019, 11(6), 634; https://doi.org/10.3390/rs11060634 - 15 Mar 2019
Abstract
The trade-off between spatial and temporal resolutions has led to the disaggregation of remotely sensed land surface temperatures (LSTs) for better applications. The window used for regression is one of the primary factors affecting the disaggregation accuracy. Global window strategies (GWSs) and local [...] Read more.
The trade-off between spatial and temporal resolutions has led to the disaggregation of remotely sensed land surface temperatures (LSTs) for better applications. The window used for regression is one of the primary factors affecting the disaggregation accuracy. Global window strategies (GWSs) and local window strategies (LWSs) have been widely used and discussed, while object-based window strategies (OWSs) have rarely been considered. Therefore, this study presents an OWS based on a segmentation algorithm and provides a basis for selecting an optimal window size balancing both accuracy and efficiency. The OWS is tested with Landsat 8 data and simulated data via the “aggregation-then-disaggregation” strategy, and compared with the GWS and LWS. Results tested with the Landsat 8 data indicate that the proposed OWS can accurately and efficiently generate high-resolution LSTs. In comparison to the GWS, the OWS improves the mean accuracy by 0.19 K at different downscaling ratios, in particular by 0.30 K over urban areas; compared with the LWS, the OWS performs better in most cases but performs slightly worse due to the increasing downscaling ratio in some cases. Results tested with the simulated data indicate that the OWS is always superior to both GWS and LWS regardless of the downscaling ratios, and the OWS improves the mean accuracy by 0.44 K and 0.19 K in comparison to the GWS and LWS, respectively. These findings suggest the potential ability of the OWS to generate super-high-resolution LSTs over heterogeneous regions when the pixels within the object-based windows derived via segmentation algorithms are more homogenous. Full article
(This article belongs to the Special Issue Object Based Image Analysis for Remote Sensing)
Show Figures

Graphical abstract

Open AccessArticle
Fully Convolutional Networks and Geographic Object-Based Image Analysis for the Classification of VHR Imagery
Remote Sens. 2019, 11(5), 597; https://doi.org/10.3390/rs11050597 - 12 Mar 2019
Cited by 5
Abstract
Land cover Classified maps obtained from deep learning methods such as Convolutional neural networks (CNNs) and fully convolutional networks (FCNs) usually have high classification accuracy but with the detailed structures of objects lost or smoothed. In this work, we develop a methodology based [...] Read more.
Land cover Classified maps obtained from deep learning methods such as Convolutional neural networks (CNNs) and fully convolutional networks (FCNs) usually have high classification accuracy but with the detailed structures of objects lost or smoothed. In this work, we develop a methodology based on fully convolutional networks (FCN) that is trained in an end-to-end fashion using aerial RGB images only as input. Skip connections are introduced into the FCN architecture to recover high spatial details from the lower convolutional layers. The experiments are conducted on the city of Goma in the Democratic Republic of Congo. We compare the results to a state-of-the art approach based on a semi-automatic Geographic object image-based analysis (GEOBIA) processing chain. State-of-the art classification accuracies are obtained by both methods whereby FCN and the best baseline method have an overall accuracy of 91.3% and 89.5% respectively. The maps have good visual quality and the use of an FCN skip architecture minimizes the rounded edges that is characteristic of FCN maps. Additional experiments are done to refine FCN classified maps using segments obtained from GEOBIA generated at different scale and minimum segment size. High OA of up to 91.5% is achieved accompanied with an improved edge delineation in the FCN maps, and future work will involve explicitly incorporating boundary information from the GEOBIA segmentation into the FCN pipeline in an end-to-end fashion. Finally, we observe that FCN has a lower computational cost than the standard patch-based CNN approach especially at inference. Full article
(This article belongs to the Special Issue Object Based Image Analysis for Remote Sensing)
Show Figures

Graphical abstract

Back to TopTop