Advances in Object-Based Image Analysis and Deep Learning

Earth observations from various platforms, satellite, airborne, ground-based, and citizen observatories, provide excellent opportunities to detect objects of interest and changes on the Earth's surface at different scales. Integrating data from various sensors, e.g., multispectral, hyperspectral, synthetic-aperture radar (SAR), and LIDAR, provides complementary and comprehensive information for monitoring Earth and understanding changes over time. A new era of unmanned aerial vehicle (UAV) applications is now possible in domains such as surveillance, disaster relief, urban planning, and urban management. Traditional remote sensing (RS) methods are inadequate for processing enormous amounts of collected data; there is a struggle, therefore, to extract useful information from highly diverse and complex RS datasets. Powerful methods and technologies are required when automatically extracting reliable and timely information.

Object-based image analysis (OBIA) groups pixels into meaningful homogenous objects and are then characterized and classified based on spectral, spatial, and contextual information. OBIA offers a practical method for handling information extraction at various sizes. The most recent advancement in AI and computer vision, deep learning (DL), provides a new perspective on feature representations and learning. A number of DL techniques have been proven to be highly effective in both computer vision and RS. Examples of these techniques include Convolutional Neural Networks (CNNs), transfer learning, domain adaptation, semantic segmentation networks, object detection models, and methods for learning with fewer labels. Advancements in both OBIA and DL, and a combination of OBIA with DL have the potential to increase the accuracy of practical applications to ground-breaking performances.

For this Special Issue, we encourage papers that present the most recent developments in object-based image analysis and deep learning for processing and analyzing remotely sensed imagery, such as aerial photography, multispectral and hyperspectral imagery, SAR, and LIDAR. This scope includes but is not limited to the following:

- Object-based image analysis,

- Semantic segmentation,

- Land cover and land use classification,

- Object-based change detection,

- Object detection and classification,

- Zero- and few-shot learning in RS imagery,

- Pre-processing techniques, e.g., image co-registration, cross-calibration, and image enhancement,

- Deep learning-based sensor fusion,

- Cloud computing and Big Earth Data in DL and OBIA,

- Edge processing for (near) real-time analysis,

- Applications of DL and OBIA, e.g., in maritime surveillance or disaster relief.

Dr. Arta Dilo
Dr. Wyke Pereboom- Huizinga
Dr. Judith Dijk
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 submissions that pass pre-check are 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 2700 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.

• deep learning
• machine learning
• object-based image analysis
• semantic segmentation
• classification
• object detection
• remote sensing
• sensor fusion

• Advances in Object-Based Image Analysis—Linked with Computer Vision and Machine Learning in Remote Sensing (10 articles)