Remote Sensing Satellite Image Applications and Multimodal Data Mining

Dear Colleagues, 

With the advent of remote sensing big data, petascale multi-source analysis-ready satellite image datasets from diverse sensors, including optical, hyperspectral, LiDAR, and synthetic aperture radar (SAR), are now widely available. The integration of these heterogeneous data sources provides multimodal data and complementary perspectives on various remote sensing applications, including environmental monitoring, land cover change, vegetation, disasters, city development, and industrial site assessment. Coupled with the booming advances in deep learning, this has driven a significant surge in demand for large-scale multimodal spatiotemporal data analysis and mining to support regional or even global remote sensing applications. Nevertheless, the data heterogeneity across different sensors introduces significant complexity among different data modalities, as the variations in spatial, spectral, and temporal resolution complicate alignment and fusion. The complexity of multi-source multimodal data also presents significant challenges in multimodal remote sensing data mining, which is spurring a wave of innovative research efforts.

Growing research efforts have focused on deep learning-based multimodal feature retrieval and mining with more sophisticated models for various remote sensing applications. Some new network architectures, such as GNN, RNN, or transformer-based models, are designed in feature-based retrieval or cross-modal matching to process and capture relevant features from each image modality simultaneously. However, these approaches typically rely solely on visual features (such as texture, shape, or spectral characteristics) extracted from different image modalities without taking semantic-rich textual information into account. This makes the data mining and retrieval more focused on basic pixel-level features, which might not always be sufficient to fully capture the semantic meaning of a remote sensing scene. Accordingly, the textual annotations, descriptions, or metadata with rich semantic information are increasingly combined with image data modalities to explore richer semantic understanding across various remote sensing applications. Meanwhile, the complex preprocessing and semantic alignment between image and textual modalities may present a challenge, but are essential for effective data mining from these satellite imageries. Moreover, the processing of heterogeneous multimodal datasets with higher dimensionality and covering larger scales could also lead to remarkable increases in computational demands and often requires more advanced parallel computing techniques and efficient algorithms.

With these issues in mind, it is time to present the current state-of-the-art theoretical, methodological, and application research on multimodal big data analysis methods for spatiotemporal remote sensing applications. This Special Issue invites articles that are related to the topic of remote sensing and geospatial big data analytics. We welcome high-quality contributions proposing solutions and approaches in the domain of, but not limited to, the following topics:

• Spatiotemporal remote sensing data analysis for environmental and ecosystem monitoring;
• Multimodal spatiotemporal data analysis for land cover change and forest and vegetation monitoring;
• Multimodal spatiotemporal data mining for city development and industrial site assessment;
• Multimodal remote sensing data mining methods aided with textual features;
• Textual feature-aided multimodal retrieval in spatiotemporal remote sensing big data;
• High-performance multi-modal remote sensing data mining accelerated with GPUs or DCUs;
• Spatiotemporal change detection with multimodal remote sensing images;
• Multimodal spatial–temporal fusion of remote sensing images.

Dr. Yan Ma
Guest Editor

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.

• spatiotemporal remote sensing big data
• multimodal data analysis on spatiotemporal remote sensing data
• multimodal image data mining with textual features
• multimodal remote sensing data retrieval and managing methods
• multimodal remote sensing data retrieval aided with textual features
• high-performance multi-modal data mining
• environment and ecosystem