GEO DATA

Data Articles

Extending Semantic Interpretation and Visual Understanding of SAR Data via Generative Translation Models: Min Kim, Kyungil Lee, Seonyoung Park; GEO DATA. 2026;8(1):44-54. Published online March 31, 2026; DOI: https://doi.org/10.22761/GD.2026.0004

406 View
17 Download

Abstract PDF: Synthetic aperture radar (SAR) offers critical all-weather observation capabilities, yet its interpretation remains challenging due to inherent speckle noise and non-intuitive scattering characteristics. Consequently, directly applying vision-language models (VLMs) trained on natural images to the SAR domain is limited by significant modality gaps and the scarcity of high-quality SAR-text datasets. To overcome these challenges, this study proposes a two-stage framework that leverages SAR-to-optical translation to bridge the domain gap. First, we introduce a conditional Brownian Bridge Diffusion Model integrated with a SAR feature guidance module. This approach transforms SAR images into optical-like representations while preserving structural fidelity, thereby addressing the geometric distortions and hallucinations common in generative adversarial network (GAN)-based methods. Second, the translated images are analyzed by a domain-adapted VLM, utilizing the GeoRSCLIP visual encoder and a LoRA-tuned LLaVA model to generate precise semantic captions. Experimental results using Sentinel-1 and Sentinel-2 datasets demonstrate that the proposed translation model outperforms existing GAN models in terms of PSNR and SSIM. Furthermore, the framework achieves significant improvements in captioning metrics, including BLEU, ROUGE-L, and BERT-Score, compared to direct SAR interpretation. This study validates that high-fidelity modality translation can effectively extend the reasoning capabilities of pre-trained VLMs to the SAR domain without requiring extensive SAR-specific annotations.

GeoAI Dataset for Building Change Detection from CAS500-1: DongHyuk Jin, Ji Sang Park, Yuna Jang, Seungha Lee, Junhwa Chi; GEO DATA. 2025;7(3):216-229. Published online September 2, 2025; DOI: https://doi.org/10.22761/GD.2025.0010

1,617 View
125 Download

Abstract PDF: Building change detection (BCD) is essential for urban planning, disaster response, and environmental monitoring. Existing datasets, such as WHU-CD, LEVIR-CD, and SYSUCD, as well as overseas satellite imagery datasets, have limitations in reflecting South Korea’s distinctive high-density urban morphology. Domestic aerial datasets, including AI-based national park change monitoring and urban BCD datasets, also suffer from varying sensor resolutions and limited applicability to satellite-based monitoring. To overcome these challenges, we introduce a high-resolution BCD dataset based on Compact Advanced Satellite 500-1 (CAS500-1), launched in 2021. This dataset utilized 0.5 m/pixel satellite imagery collected from 2021 to 2024 across diverse regions, including Seoul, Busan, and Daejeon. The dataset construction involved precise image alignment, RGB band synthesis, histogram stretching for contrast enhancement, and polygon-based labeling to identify urban changes. Additionally, format conversions were applied to ensure deep learning compatibility, generating 3,803 paired image patches (256×256 pixels). To validate the dataset, we employed ChangeMamba, a model integrating a visual state space model for spatiotemporal change detection. The model trained on our dataset achieved an F1 score of 0.79 and an intersection over union (IoU) of 0.66, outperforming models trained on different BCD datasets in cross-dataset tests, where the latter scored significantly lower (e.g., F1, 0.18 and IoU, 0.10 for the model trained on LEVIR-CD). Moreover, our dataset demonstrated improved consistency over domestic aerial datasets, which often suffer from heterogeneous resolutions and limited scalability for long-term monitoring. By leveraging high-resolution satellite imagery, our dataset provides a robust alternative to both overseas and domestic aerial datasets, enabling precise urban monitoring with reduced reliance on aerial surveys. It serves as a critical resource for advancing BCD research in South Korea. Future research will focus on automating the labeling process, expanding temporal coverage, and improving model performance through reinforcement learning on pre-trained models using this dataset.

Analysis of Changes in Vegetation Cover Using Satellite Images of Sorok Island in Republic of Korea: Hyungjin Cho, Seungnam Jin, Jong-Seo Won, Youngjun Park; GEO DATA. 2025;7(2):62-69. Published online June 12, 2025; DOI: https://doi.org/10.22761/GD.2025.0002

1,539 View
46 Download
1 Citations

Abstract PDF

This study aims to compare and analyze the past and present vegetation changes on Sorok Island using satellite imagery to understand the current development patterns that damage the island's unique ecological and environmental characteristics. The findings will serve as fundamental data for establishing sustainable management and conservation plans. Sorok Island, located in Goheung-gun, South Korea, has a historical background as a Hansen's disease patient isolation site. Since 1916, limited human intervention has shaped its unique ecosystem. This study uses Sentinel-2 satellite images from 2016 and 2024 to calculate normalized difference vegetation index (NDVI) and analyze vegetation vitality and change rates. Results indicate that NDVI values in 2024 have generally increased compared to 2016, and vegetation vitality has also improved. However, NDVI values for forest areas and inland wetlands have decreased, attributed to forest development and expansion of cultivated land. Future research should include long-term vegetation change analysis and comprehensive studies considering various factors.

Citations

Citations to this article as recorded by

First Avifaunal Inventory of Sorok Island, South Korea
Sang-Yeon Lee
GEO DATA.2025; 7(3): 137. CrossRef

KOMPSAT-3/3A Image-text Dataset for Training Large Multimodal Models: Han Oh, Dong-Bin Shin, Dae-Won Chung; GEO DATA. 2025;7(1):27-35. Published online March 19, 2025; DOI: https://doi.org/10.22761/GD.2025.0003

2,191 View
96 Download
1 Citations

Abstract PDF

This study aims to improve the accuracy and interpretability of large multimodal models (LMMs) specialized in satellite image analysis by constructing an image-text dataset based on KOMPSAT-3/3A imagery and presenting the results of training using this dataset. Conventional LMMs are primarily trained on general images, limiting their ability to effectively interpret the specific characteristics of satellite imagery, such as spectral bands, spatial resolution, and viewing angles. To address this limitation, we developed an image-text dataset, divided into pretraining and finetuning stages, based on the existing KOMPSAT object detection dataset. The pretraining dataset consists of captions summarizing the overall theme and key information of each image. The fine-tuning dataset integrates metadata -including acquisition time, sensor type, and coordinates- with detailed object detection labels to generate six types of question-answer pairs: detailed descriptions, conversations with varying answer lengths, bounding box identification, multiple choice questions, and complex reasoning. This structured dataset enables the model to learn not only the general context of satellite images but also fine-grained details such as object quantity, location, and geographic attributes. Training with the new KOMPSAT-based dataset significantly improved the model’s accuracy in recognizing regional information and object characteristics in satellite imagery. Finetuned models achieved substantially higher accuracy than previous models, surpassing even the GPT-4o model and demonstrating the effectiveness of a domain-specific dataset. The findings of this study are expected to contribute to various remote sensing applications, including automated satellite image analysis, change detection, and object detection.

Citations

Citations to this article as recorded by

KFGOD: A Fine-Grained Object Detection Dataset in KOMPSAT Satellite Imagery
Dong Ho Lee, Ji Hun Hong, Hyun Woo Seo, Han Oh
Remote Sensing.2025; 17(22): 3774. CrossRef

Original Papers

The Cheonji Lake GeoAI Dataset based in Optical Satellite Imagery: Landsat-5/-7/-8 and Sentinel-2: Eu-Ru Lee, Ha-Seong Lee, Sun-Cheon Park, Hyung-Sup Jung; GEO DATA. 2024;6(1):14-23. Published online March 28, 2024; DOI: https://doi.org/10.22761/GD.2023.0055

4,213 View
88 Download
1 Citations

Abstract PDF

The variations in the water area and water level of Cheonji, the caldera lake of Baekdu Mountain, serve as reliable indicators of volcanic precursors. However, the geographical and spatial features of Baekdusan make it impossible to directly observe the water area and water level. Therefore, it is crucial to rely on remote sensing data for monitoring purposes. Optical satellite imagery employs different spectral bands to accurately delineate the boundaries between water bodies and non-water bodies. Conventional methods for classifying water bodies using optical satellite images are significantly influenced by the surrounding environment, including factors like terrain and shadows. As a result, these methods often misclassify the boundaries. To address these limitations, deep learning techniques have been employed in recent times. Hence, this study aimed to create an AI dataset using Landsat-5/-7/-8 and Sentinel-2 optical satellite images to accurately detect the water body area and water level of Cheonji lake. By utilizing deep learning methods on the dataset, it is reasonable to consistently observe the area and level of water in Cheonji lake. Furthermore, by integrating additional volcanic precursor monitoring factors, a more accurate volcano monitoring system can be established.

Citations

Citations to this article as recorded by

Performance Comparison of Water Body Detection from Sentinel-1 SAR and Sentinel-2 Optical Imagery Using Attention U-Net Model
Il-Hoon Choi, Eu-Ru Lee, Hyung-Sup Jung
Korean Journal of Remote Sensing.2024; 40(5-1): 507. CrossRef

GeoAI Dataset for Industrial Park and Quarry Classification from KOMPSAT-3/3A Optical Satellite Imagery: Che-Won Park, Hyung-Sup Jung, Won-Jin Lee, Kwang-Jae Lee, Kwan-Young Oh, Jae-Young Chang, Moung-jin Lee, Geun-Hyouk Han, Il-Hoon Choi; GEO DATA. 2023;5(4):238-243. Published online December 28, 2023; DOI: https://doi.org/10.22761/GD.2023.0052

3,062 View
96 Download
4 Citations

Abstract PDF

Air pollution is a serious problem in the world, and it is necessary to monitor air pollution emission sources in other neighboring countries to respond to the problem of air pollution spreading across borders. In this study, we utilized domestic and international optical images from KOMPSAT-3/3A satellites to build an AI training dataset for classifying industrial parks and quarries, which are representative sources of air pollution emissions. The data can be used to identify the distribution of air pollution emission sources located at home and abroad along with various state-of-the-art models in the image segmentation field, and is expected to contribute to the preservation of Korea’s air environment as a basis for establishing air-related policies.

Citations

Citations to this article as recorded by

Tracing the evolution of urban flood vulnerability: assessing the impact of subsurface drainage upgrades in Seoul using deep learning
Junhyeok Jung, Euru Lee, Hyung-Sup Jung
Geosciences Journal.2026;[Epub] CrossRef
GeoAI Dataset for Industrial Park Segmentation from Sentinel-2 Satellite Imagery and GEMS
Sung-Hyun Gong, Hyung-Sup Jung, Geun-han Kim, Geun-Hyouk Han, Il-Hoon Choi, Jin-Sung Hong
GEO DATA.2025; 7(1): 36. CrossRef
Semantic Segmentation of Urbanized Areas Using Multi-Encoder U-Net Based on Multi-Modal Data
Sung-Hyun Gong, Hyung-Sup Jung, Geun-Han Kim, Geun-Hyouk Han, Il-Hoon Choi, Jin-Sung Hong
Korean Journal of Remote Sensing.2025; 41(2): 461. CrossRef
Performance Comparison of Water Body Detection from Sentinel-1 SAR and Sentinel-2 Optical Imagery Using Attention U-Net Model
Il-Hoon Choi, Eu-Ru Lee, Hyung-Sup Jung
Korean Journal of Remote Sensing.2024; 40(5-1): 507. CrossRef

GeoAI Dataset for Training Deep Learning-Based Optical Satellite Image Matching Model: Jin-Woo Yu, Che-Won Park, Hyung-Sup Jung; GEO DATA. 2023;5(4):244-250. Published online December 28, 2023; DOI: https://doi.org/10.22761/GD.2023.0048

2,988 View
97 Download
3 Citations

Abstract PDF

Satellite imagery is being used to monitor the Earth, as it allows for the continuous provision of multi-temporal observations with consistent quality. To analyze time series remote sensing data with high accuracy, the process of image registration must be conducted beforehand. Image registration techniques are mainly divided into region-based registration and feature-based registration, and both techniques extract the same points based on the similarity of spectral characteristics and object shapes between master and slave images. In addition, recently, deep learning-based siamese neural network and convolutional neural network models have been utilized to match images. This has high performance compared to previous non-deep learning algorithms, but a very large amount of data is required to train a deep learning-based image registration model. In this study, we aim to generate a dataset for training a deep learning-based optical image registration model. To build the data, we acquired Satellite Side-Looking (S2Looking) data, an open dataset, and performed preprocessing and data augmentation on the data to create input data. After that, we added offsets to the X and Y directions between the master and slave images to create label data. The preprocessed input data and labeled data were used to build a dataset suitable for image registration. The data is expected to be useful for training deep learning-based satellite image registration models.

Citations

Citations to this article as recorded by

Center Point Matching Network Based on Multi-Coverage Templates for High-Resolution Satellite Orthorectification
Jaewon Hur, Junseo Baek, Doochun Seo, Changno Lee, Jaehong Oh, Jaewan Choi, Youkyung Han
KSCE Journal of Civil Engineering.2026; : 100584. CrossRef
WorldView-3 Super-resolution Training Dataset Construction Using the MTF-GLP Method
Nayeon Kim, ChanHee Jeong, Gibeom Nam, Kimook Kang, Euiho Hwang, DongHyeon Yoon
GEO DATA.2025; 7(4): 835. CrossRef
Performance Comparison of Water Body Detection from Sentinel-1 SAR and Sentinel-2 Optical Imagery Using Attention U-Net Model
Il-Hoon Choi, Eu-Ru Lee, Hyung-Sup Jung
Korean Journal of Remote Sensing.2024; 40(5-1): 507. CrossRef

Articles

AI Dataset for Road Detection using KOMPSAT Images: Hoonhee Lee, Han Oh; GEO DATA. 2022;4(1):43-48. Published online March 31, 2022; DOI: https://doi.org/10.22761/DJ2022.4.1.005

3,575 View
61 Download
1 Citations

Abstract PDF

Information on shape and type of road present in an optical image of satellite is useful for digital mapping and monitoring of road changes. Processing and structuring optical image data collected from payloads mounted on KOMPSAT 3 and 3A can accelerate the development of road detection algorithms and the extraction of road information using them. In particular, if it is built with a learning dataset for AI (Artificial Intelligence) prepared to apply deep learning technology, the latest artificial intelligence technology in the field of computer science can be spun off to the field of satellite image-based road detection to attempt a wide range of analysis. Korea Aerospace Research Institute constructed an image dataset for AI learning using satellite optical images with Korean companies, and this paper explains the type and size of datasets along with examples of the use of the dataset. The established data can be used through the website, aihub.or.kr.

Citations

Citations to this article as recorded by

Strategies for Reducing Ground Station Operational Costs of Optical Satellite Constellations
Keon-Woo Koo, Jaekwang Kim, Dohyun Heo, Minkyo Park, Hae-Dong Kim, Dong-Ho Yun
International Journal of Aeronautical and Space Sciences.2026;[Epub] CrossRef

AI Training Dataset for Cloud Detection of KOMPSAT Images: Bo-Ram Kim, Han Oh; GEO DATA. 2020;2(2):56-62. Published online December 30, 2020; DOI: https://doi.org/10.22761/DJ2020.2.2.008

2,698 View
73 Download
1 Citations

Abstract PDF

Clouds that appear inevitably when acquiring optical satellite images hinder the interpretation of surface information, so removing them is a crucial procedure to increase the utilization of satellite images. Currently, for KOMPSAT (Korea Multi-purpose Satellite) images, only the cloud amount by visual measurement is proved for the entire scene and detailed cloud masks are not provided. Since cloud detection is a time-consuming task, we built a cloud dataset for KOMPSAT images so as to develop an algorithm that expedites the task with state-of-the-art artificial intelligent techniques. In the dataset, satellite images were selected from various regions considering that clouds have different characteristics depending on the region, and masks were classified into thin clouds, thick clouds, cloud shadows, and clear sky. The size of dataset is over 4,000 image/mask pairs by an image size of 1000x1000 and one of the largest among publicly available cloud datasets, as of this writing. The dataset is built by a government AI (artificial intelligent) training dataset building program and will be available through the website, aihub.or.kr.

Citations

Citations to this article as recorded by

Cloud Detection Using a UNet3+ Model with a Hybrid Swin Transformer and EfficientNet (UNet3+STE) for Very-High-Resolution Satellite Imagery
Jaewan Choi, Doochun Seo, Jinha Jung, Youkyung Han, Jaehong Oh, Changno Lee
Remote Sensing.2024; 16(20): 3880. CrossRef

First
Prev
Page of 1
Next
Last

Search