About this Research Topic
Disaster has significantly threatened land resources over the past few decades due to natural factors such as flood hazards and human factors including growing populations. To monitor the occurrence of disasters and carry out disaster rescues, it is vital to achieve a disaster emergency response. Multi-modal remote sensing data have become one of the most useful resources available for disaster response, such as geographic information data, space-borne data, unmanned aerial vehicle data, ground spectrum, field photos, temperature, humidity, etc. Because these multimodal data contain different data structures, extracting valuable information from them is hard to support emergency response decision-making. Therefore, it is meaningful to work on multi-modal remote sensing disaster response.
However, some limitations need to be considered: 1) these multi-modal data are captured with various geographic references, so it is difficult to effectively collect multi-modal data in a consistent spatial reference; 2) remote sensing data with a certain model (e.g., optical data) may be missing since they suffer from extreme weather conditions such as cloudy weather that makes disaster areas unseen; 3) with limited resources, it is difficult to rapidly obtain valuable information from multi-modal data to support disaster response decision making. A balance between accuracy and efficiency is required to provide real-time but accurate responses; 4) damage assessment for geo-objects is necessary to acquire the situation of a disaster. Yet, it is difficult to accurately provide the damage assessment results.
Therefore, the four key problems of disaster emergency response are: (1) How to effectively organize the multi-model data and achieve an accurate image registration and fusion; (2) How to carry out multi-model image translation on the transformation of available data (e.g., SAR) to unavailable data (optical images); (3) How to rapidly and accurately learn valuable disaster response information from multi-modal data; (4) How to assess the damage level of geo-objects to reveal the situation of a disaster.
This Research Topic would explore cross-disciplinary approaches, methodologies, and applications of disaster response algorithms, tools, procedures, and models that can be incorporated into an accurate and efficient system. This Research Topic calls for researchers in remote sensing, damage assessment, sustainable development, public policy, and other disciplines and cross-disciplinary fields. Topics of interest include, but are not limited to:
• Multi-modal data registration, fusion, and quality enhancement;
• Multi-modal data information extraction for disaster response;
• Semantic segmentation and object detection in remote sensing;
• Multi-modal remote sensing data change detection;
• Real-time processing of multi-modal remote sensing data;
• Multi-modal data damage assessment;
• Deep learning theory and its applications in disaster emergency response.
However, some limitations need to be considered: 1) these multi-modal data are captured with various geographic references, so it is difficult to effectively collect multi-modal data in a consistent spatial reference; 2) remote sensing data with a certain model (e.g., optical data) may be missing since they suffer from extreme weather conditions such as cloudy weather that makes disaster areas unseen; 3) with limited resources, it is difficult to rapidly obtain valuable information from multi-modal data to support disaster response decision making. A balance between accuracy and efficiency is required to provide real-time but accurate responses; 4) damage assessment for geo-objects is necessary to acquire the situation of a disaster. Yet, it is difficult to accurately provide the damage assessment results.
Therefore, the four key problems of disaster emergency response are: (1) How to effectively organize the multi-model data and achieve an accurate image registration and fusion; (2) How to carry out multi-model image translation on the transformation of available data (e.g., SAR) to unavailable data (optical images); (3) How to rapidly and accurately learn valuable disaster response information from multi-modal data; (4) How to assess the damage level of geo-objects to reveal the situation of a disaster.
This Research Topic would explore cross-disciplinary approaches, methodologies, and applications of disaster response algorithms, tools, procedures, and models that can be incorporated into an accurate and efficient system. This Research Topic calls for researchers in remote sensing, damage assessment, sustainable development, public policy, and other disciplines and cross-disciplinary fields. Topics of interest include, but are not limited to:
• Multi-modal data registration, fusion, and quality enhancement;
• Multi-modal data information extraction for disaster response;
• Semantic segmentation and object detection in remote sensing;
• Multi-modal remote sensing data change detection;
• Real-time processing of multi-modal remote sensing data;
• Multi-modal data damage assessment;
• Deep learning theory and its applications in disaster emergency response.
Keywords: Multi-model remote sensing, disaster response, damage assessment, real-time processing, image registration, image fusion, deep learning
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
