Special Issue "Deep Learning Algorithms for Land Use Change Detection"

A special issue of Land (ISSN 2073-445X).

Deadline for manuscript submissions: closed (1 April 2021).

Special Issue Editors

Special Issue Information

Dear Colleagues,

The rapid growth of the human population in recent decades has increased the demand for residential lands, natural resources, and basic services. The increased shift of population from rural to urban areas leads to rapid urbanization, and this has resulted in huge increase in the amount of built-up areas. The identification and management of land-use changes related to these trajectories have become imperative for meeting the needs of growing populations. Remote sensing techniques can be used to monitor changes in land use and land cover. The data used includes images of land use, weather and soil conditions, and more. The massive volume of land data demands an effective big data analytics process, so that unknown and complex patterns can be discovered. Most of the existing research works make use of aerial and space-borne scene classification to determine land-use changes. Here, the images are labelled under predefined conditions such as vegetation, housing, service facilities, etc. The changes in land use and land cover are documented with the help of remote sensing techniques. Advances in technology have enriched the growth of remote sensing, but at the same time traditional approaches have become less appropriate for the detection of land-use and land-cover changes. In order to provide meaningful analysis of the huge amounts of data, big data analytics is increasingly used because of its ability to reveal patterns in complex situations. A major challenge for big data analytics in recent times has been the increase in data volumes. This increases the requirement for the development and use of advanced learning algorithms to find and interpret dynamic changes in land use.

Deep learning (DL) algorithms can efficiently handle huge volumes of data, and can classify them with increased accuracy. Examples of important land-use detection include DL algorithms like convolutional neural networks (CNNs), recurrent neural networks (RNNs), and deep belief networks (DBNs). The input data passes across the multiple hidden layers through which the patterns are recognized and classified. Data collection, feature extraction, pattern recognition, data training, and classification are key features of these algorithms. 

Diversity in land use brings out the heterogeneity in the datasets. DL algorithms classify the datasets, which are then fed into low-level and high-level layers based on the level of data complexity. This enables algorithms to easily extract the core features of the datasets, which enhances the quality of image classification methods and can improve pre-processing and segmentation tasks, handle high-dimensional data, and perform well with limited datasets.

This Special Issue on “Deep Learning Algorithms for Land-Use Change Detection” encourages researchers and data scientists from various research backgrounds to present their novel ideas and algorithms for the detection of land-use change detection with DL techniques.

The topics of interest include but are not limited to the following:

  • Convolutional neural networks (CNNs) for mapping land-cover changes;
  • Deep learning (DL) methods for understanding the impact of growth in population and land usage;
  • Modeling a deep belief network (DBN) to determine land-use change in urban areas;
  • Big data for land-use change detection: challenges and opportunities;
  • Quality assessment of land cover big data;
  • Deep learning (DL) algorithms to evaluate landscape changes;
  • Surveys on existing deep learning (DL) techniques to classify land use;
  • Meta-analysis for core feature extraction of land cover in developed countries;
  • Deep learning for resolving the challenges for land-use and land-cover change;
  • Applications of deep learning and big data analytics in land-use and land-cover change detection;
  • Case studies on global land cover monitoring and big data challenges;
  • The role of convolutional neural networks (CNNs) in documenting land degradation;
  • Recent advancements in deep learning (DL) for the quality assessment of soil, water, and air due to land-use change;
  • Deep learning approaches to understanding the influences of population growth on land-use change;
  • Land-use change classification using deep learning approaches;
  • Deep learning and big data analytics for land detection and knowledge discovery.

Dr. Gunasekaran Manogaran
Dr. Hassan Qudrat-Ullah
Dr. Qin Xin
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. Land 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 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

  • big data analytics
  • deep learning (DL)
  • deep belief network (DBN)
  • convolutional neural network (CNN)
  • landscape changes
  • land degradation
  • land-use change classification

Published Papers (2 papers)

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

Research

Open AccessEditor’s ChoiceArticle
Exploring the Sensitivity of Recurrent Neural Network Models for Forecasting Land Cover Change
Land 2021, 10(3), 282; https://0-doi-org.brum.beds.ac.uk/10.3390/land10030282 - 10 Mar 2021
Viewed by 271
Abstract
Recurrent Neural Networks (RNNs), including Long Short-Term Memory (LSTM) architectures, have obtained successful outcomes in timeseries analysis tasks. While RNNs demonstrated favourable performance for Land Cover (LC) change analyses, few studies have explored or quantified the geospatial data characteristics required to utilize this [...] Read more.
Recurrent Neural Networks (RNNs), including Long Short-Term Memory (LSTM) architectures, have obtained successful outcomes in timeseries analysis tasks. While RNNs demonstrated favourable performance for Land Cover (LC) change analyses, few studies have explored or quantified the geospatial data characteristics required to utilize this method. Likewise, many studies utilize overall measures of accuracy rather than metrics accounting for the slow or sparse changes of LC that are typically observed. Therefore, the main objective of this study is to evaluate the performance of LSTM models for forecasting LC changes by conducting a sensitivity analysis involving hypothetical and real-world datasets. The intent of this assessment is to explore the implications of varying temporal resolutions and LC classes. Additionally, changing these input data characteristics impacts the number of timesteps and LC change rates provided to the respective models. Kappa variants are selected to explore the capacity of LSTM models for forecasting transitions or persistence of LC. Results demonstrate the adverse effects of coarser temporal resolutions and high LC class cardinality on method performance, despite method optimization techniques applied. This study suggests various characteristics of geospatial datasets that should be present before considering LSTM methods for LC change forecasting. Full article
(This article belongs to the Special Issue Deep Learning Algorithms for Land Use Change Detection)
Show Figures

Figure 1

Open AccessArticle
Generative Learning for Postprocessing Semantic Segmentation Predictions: A Lightweight Conditional Generative Adversarial Network Based on Pix2pix to Improve the Extraction of Road Surface Areas
Land 2021, 10(1), 79; https://0-doi-org.brum.beds.ac.uk/10.3390/land10010079 - 16 Jan 2021
Viewed by 797
Abstract
Remote sensing experts have been actively using deep neural networks to solve extraction tasks in high-resolution aerial imagery by means of supervised semantic segmentation operations. However, the extraction operation is imperfect, due to the complex nature of geospatial objects, limitations of sensing resolution, [...] Read more.
Remote sensing experts have been actively using deep neural networks to solve extraction tasks in high-resolution aerial imagery by means of supervised semantic segmentation operations. However, the extraction operation is imperfect, due to the complex nature of geospatial objects, limitations of sensing resolution, or occlusions present in the scenes. In this work, we tackle the challenge of postprocessing semantic segmentation predictions of road surface areas obtained with a state-of-the-art segmentation model and present a technique based on generative learning and image-to-image translations concepts to improve these initial segmentation predictions. The proposed model is a conditional Generative Adversarial Network based on Pix2pix, heavily modified for computational efficiency (92.4% decrease in the number of parameters in the generator network and 61.3% decrease in the discriminator network). The model is trained to learn the distribution of the road network present in official cartography, using a novel dataset containing 6784 tiles of 256 × 256 pixels in size, covering representative areas of Spain. Afterwards, we conduct a metrical comparison using the Intersection over Union (IoU) score (measuring the ratio between the overlap and union areas) on a novel testing set containing 1696 tiles (unseen during training) and observe a maximum increase of 11.6% in the IoU score (from 0.6726 to 0.7515). In the end, we conduct a qualitative comparison to visually assess the effectiveness of the technique and observe great improvements with respect to the initial semantic segmentation predictions. Full article
(This article belongs to the Special Issue Deep Learning Algorithms for Land Use Change Detection)
Show Figures

Figure 1

Back to TopTop