Jeonbuk National University (JBNU) Department of Environment and Energy Convergence, AI Remote Sensing Laboratory master's students Han Do-hee, Kim Su-hwan, and Lee Ye-jin (supervisor Professor Yeom Jong-min) received the Excellence Award at the nationwide "2025 Space Hackathon" for research combining satellite data and artificial intelligence.
 

The competition was organized to generate creative ideas using satellite data and artificial intelligence technologies. It was held in three categories: Satellite Big Data Utilization; Satellite Technology-Based Business Models; and Small-Satellite Mission Ideas. A team composed of JBNU students participated in the "Satellite Big Data Utilization Ideas" category and received the Excellence Award.

The title of the winning entry was "Pansharpening Technique-Applied KOMPSAT-3 Imagery-Based Building Damage Detection Tool-Integrated LLM Visual Reasoning Framework." The entry was notable for proposing a disaster-response technology that combines multi-satellite imagery and artificial intelligence.

The research team applied a pan-sharpening technique that fuses high-resolution panchromatic (PAN) images with multispectral images to enhance spatial resolution and spectral information simultaneously, and built an automated analysis system to improve building damage detection accuracy.

In particular, they linked large-scale satellite imagery analysis results with large language models (LLMs) to enable classification of damage types, summarization of damage severity, and generation of decision-support information. This design was highly evaluated because it allows rapid and intuitive information delivery during disasters.

Professor Yeom Jong-min, the team's supervisor, said, "This study, which integrates high-resolution satellite imagery, artificial intelligence, and language models, demonstrates a new direction for the use of satellite data," and added, "It is meaningful that the students developed the technical skills accumulated during the research into a practical, problem-solving idea."

The winning team said, "Based on the AI-based remote sensing research conducted in the laboratory, we aimed to propose technology that can make a practical contribution to disaster response," and added, "We will enhance the model and expand the range of satellite data to increase the feasibility of field deployment."