Research Areas
Satellite Image Analysis
Our lab utilized spectral & thermal IR imagery to investigate the Urban Heat Island (UHI) effect in metropolitan areas like Seoul and Gumi. Our research identified the correlation between Land Surface Temperature (LST) and vegetation indices (NDVI), specifically revealing "anomalous variations" where artificial green structures (e.g., green paint on factory roofs) in industrial complexes were misclassified as vegetation. We also applied fractal geometry (power law) to analyze the scaling effects of lake distributions using satellite-derived water masks.
Key Publications
Drone Remote Sensing
Our work established a methodology for the 3D digital restoration of Korean traditional architecture (Hanok) by fusing terrestrial laser scanning (LiDAR) point clouds with high-resolution photogrammetry. Furthermore, we pioneered the use of hyperspectral sensors (VNIR cameras) and spectrometers to analyze the spectral signatures of concrete structures. This allows for the automated detection of aging, carbonation, and micro-cracks that are invisible to standard RGB cameras, serving as a non-destructive testing method for infrastructure.
Key Publications
Environmental & Disaster Monitoring
The research team developed a "Smart Pole" measurement system designed for monitoring steep slopes and landslide risks. This IoT-based system integrates GNSS receivers, tilt sensors, and soil moisture sensors to detect real-time ground displacement and transmit warning data. Additionally, we applied the WRF (Weather Research and Forecasting) model to downscale global climate data, enabling precise habitat analysis for endangered species (Long-tailed Goral) vulnerable to climate change and extreme weather events.
Key Publications
Uncertainty Modeling
A significant portion of our research focuses on quantifying errors in global digital elevation models (DEMs). We rigorously verified the vertical accuracy of SRTM (Shuttle Radar Topography Mission) data using ICESat GLAS laser altimetry as ground truth. Our findings highlighted specific limitations and errors when using C-band radar DEMs for hydrologic modeling (e.g., stream network extraction, basin boundary delineation) in low-relief terrains, providing critical correction factors for water resource management.
Key Publications
Deep Learning-Based Decision (GeoAI)
Moving beyond traditional statistics, our lab integrated Multi-Layer Perceptron (MLP) artificial neural networks with Markov Chain Analysis (MCA) to predict future land cover changes. This GeoAI approach successfully modeled complex urban sprawl patterns in the Seoul metropolitan area, simulating future scenarios (e.g., for the year 2026) based on historical Landsat data training. This demonstrates the application of AI in strategic urban planning and environmental impact assessment.
Key Publications
Geointelligence and Cyber-security
Geospatial information is directly tied to digital twins, national defense, and national security. The level of realism in how geospatial information is represented can threaten the security of the nation, society, and individuals, and can put lives and property at risk. Through research on anomaly detection in imagery and improving the reliability of geospatial information, we aim to prevent tampering and forgery and ensure the integrity of geospatial data collected by UAVs and satellites, thereby building a trusted data framework for monitoring critical infrastructure.
Director's Biography
Kon Joon Bhang, Ph.D.
Professor Bhang is a leading expert in Remote Sensing and GIS. With a foundation in Civil and Environmental Engineering and advanced degrees from The Ohio State University, he has dedicated his career to bridging the gap between physical infrastructure and digital data. His research integrates Remote Sensing, GIS, and AI to solve critical environmental and urban challenges.
- 🎓 Ph.D. in Environmental Science Graduate Program (ESGP), The Ohio State University (2008)
- 🎓 M.S. in Civil and Environmental Engeering and Geodetic Sciences, The Ohio State University (2003)
- 🎓 B.S. in Civil and Environmental Engineering, Kookmin University (1998)
Geointelligence & Cyber Security
Developing a Cyber-Physical Security Framework to ensure reliability of geospatial data.
The Era of GeoAI & Convergence
Leading the integration of Deep Learning (MLP, MCA) with GIS for urban prediction. Expanded research to ecological digital twins, including Long-tailed Goral habitat modeling.
The Rise of Environmental Monitoring & Risk Analysis
Developing remote sensing and GIS approaches to track land surface dynamics and emerging hazards. Established the groundwork for later work in resilience, disaster monitoring, and decision-support mapping.
Joined Kumoh National Institute of Technology
Appointed as Professor. Established the Spatial Information Laboratory in Gumi, focusing on industrial environmental monitoring and urban heat island (UHI) effects.
Join Our Lab
We are looking for passionate individuals to join our team. Whether you are an undergraduate student just starting out, or an experienced Ph.D. researcher, we have a place for you.
Postdoctoral Researcher (GeoAI)
Key Responsibilities
- Conduct GeoAI & geospatial data reliability research.
- Process satellite/UAV/sensor data for Deep Learning.
- Write technical reports and journal papers.
- Collaborate with graduate students and support national projects.
Qualifications
- Required: Ph.D. in Geospatial Science, Remote Sensing, AI, or related field.
- Preferred: Proficiency in Python (PyTorch/TensorFlow).
- Preferred: Experience with GDAL, QGIS, or ArcGIS.
Graduate & Undergraduate Researchers
We are seeking motivated students interested in the convergence of Civil & Environmental Engineering, artificial intelligence, and ICT. You will gain hands-on experience in national R&D projects and advanced data analysis.
What you will learn:
- Handling Geospatial Data (Satellite & Drone imagery) and analyses.
- Applying AI algorithms to real-world environmental problems.
- Cyber-Physical security framework for geospatial data
- Academic writing and developing methodology.