Since the release of Meta’s Segment Anything Model (SAM), a large-scale vision transformer generation model with rapid image segmentation capabilities, several studies have been conducted to apply this technology in various fields. In this study, we aimed to investigate the applicability of SAM for water bodies detection and extraction using the QGIS Geo-SAM plugin, which enables the use of SAM with satellite imagery. The experimental data consisted of Compact Advanced Satellite 500 (CAS500)-1 images. The results obtained by applying SAM to these data were compared with manually digitized water objects, Open Street Map (OSM), and water body data from the National Geographic Information Institute (NGII)-based hydrological digital map. The mean Intersection over Union (mIoU) calculated for all features extracted using SAM and these three-comparison data were 0.7490, 0.5905, and 0.4921, respectively. For features commonly appeared or extracted in all datasets, the results were 0.9189, 0.8779, and 0.7715, respectively. Based on analysis of the spatial consistency between SAM results and other comparison data, SAM showed limitations in detecting small-scale or poorly defined streams but provided meaningful segmentation results for water body classification.

As disasters become more diverse and widespread, disaster management at the national level is increasingly important. Since satellite remote sensing technology is capable of observing a wide range of areas on a regular basis, it can be effectively utilized to monitor a massive scale of disaster conditions and preemptively respond to urgent disasters. Disasters that can be managed through satellite remote sensing technology include forest fires, drought, floods, landslides, and earthquakes. This study introduces a variety of studies using satellite remote sensing and Geographic Information System (GIS) as well as a number of actual cases utilizing the above. However, with the satellites currently in operation, there are difficulties in detecting and analyzing the disasters above due to the limitations on spatial and temporal resolutions. Recently, new measures have been developed to overcome such limitations through the development and constellation operation of microsatellites. In addition, new technologies are under development where a massive quantity of satellite images is analyzed by Artificial Intelligence (AI) technology. It is expected that temporal and spatial limitations can be addressed through satellite-developed and constellation systems in the future, which would lead to scientific disaster management through grafting with AI technology.

Microplastics have arisen as a worldwide environmental concern, becoming ubiquitous in all marine compartments, and various researches on monitoring marine microplastics are being actively conducted worldwide. Recently, application of a remote detection technology that enables large-scale real-time observation to marine plastic monitoring has been conducted overseas. However, in South Korea, there is little information linking remote detection to marine microplastics and some field studies have demonstrated remote detection of medium- and large-sized marine plastics. This study introduces research cases with remote detection of marine plastics in South Korea and overseas, investigates potential feasibility of using the remote detection technology to marine microplastic monitoring, and suggests some future works to monitor marine microplastics with the remote detection.

Semantic image segmentation techniques have recently gained widespread adoption in the field of remote sensing for tasks such as classifying surface properties and extracting specific objects. The performance of semantic image segmentation is influenced not only by the choice of deep learning model but also by the configuration of key hyperparameters, including learning rate and batch size. Among these hyperparameters, the batch size is typically set to a larger value to improve model performance. However, since the processing capacity of a typical deep learning system’s graphics processing unit (GPU) is limited, selecting an appropriate batch size is necessary. This paper investigates the impact of batch size on building detection performance in deep learning systems for semantic image segmentation using satellite and aerial imagery. For the performance analysis, representative models for semantic image segmentation, including UNet, ResUNet, DeepResUNet, and CBAM-DRUNet, were used as baseline models. Additionally, transfer learning models such as UNet-VGG19, UNet-ResNet50, and CBAM-DRUNet-VGG19 were incorporated for comparison. The training datasets used in this study included the WHU and INRIA datasets, which are commonly used for semantic image segmentation tasks, as well as the Kompsat-3A dataset. The experimental results revealed that a batch size of 2 or larger led to an improvement in F1 scores across all models and datasets. For the WHU dataset, the smallest of the datasets, the F1 score initially increased with batch size, but after reaching a certain threshold, it began to decline, except for the CBAM-DRUNet-VGG19 model. In contrast, for the INRIA dataset, which is approximately 1.5 times larger than WHU, transfer learning models maintained relatively stable F1 scores as the batch size increased, while other models showed a similar trend of increasing F1 scores followed by a decrease. In the case of the Kompsat-3A datasets, which are 4 to 5 times larger than the WHU dataset, all models showed a substantial increase in F1 score when the batch size was set to 2. Beyond this point, F1 scores stabilized without further significant improvements. In terms of training time, increasing the batch size generally resulted in reduced training time for all models. Therefore, when the training dataset is sufficiently large, setting the batch size to 2 is already sufficient to achieve significant improvements in F1 score accuracy. Furthermore, setting the batch size to a value greater than 2 may be advantageous in terms of further reducing training time, provided that the GPU capacity of the deep learning system is sufficient to handle the larger batch size.

Cerro Azul volcano, one of the most active volcanoes in the Galápagos Archipelago, exhibited significant unrest in March 2017, marked by dramatic displacement around the caldera and on the southeastern plain. This study applied differential interferometric synthetic aperture radar (DInSAR) technique to SAR image pairs acquired from the Sentinel-1A satellite’s ascending and descending nodes to measure the line-of-sight (LOS) and two-dimensional (up-down and east-west) displacements during the 2017 unrest of the volcano. Extensive concentric fringe patterns were observed around the caldera and on the southeastern plain of Cerro Azul volcano, indicating LOS displacements away from and toward the satellite, respectively. Two-dimensional displacements revealed that the volcanic body around the caldera contracted, while the southeastern plain uplifted and expanded. To further characterize magmatic activity during the unrest, we inverted LOS displacements from both ascending and descending observations using Mogi and Sill source models. The surface displacements simulated from the optimal magma source parameters obtained through inversion showed a normalized root mean square error of less than 5% compared to the DInSAR-observed displacements, confirming the reliability of the inversion results. The Mogi source was located approximately 4.7 km beneath the northeastern caldera rim of Cerro Azul and exhibited a volume change of –31 × 106 m3. The Sill source, characterized by a long rectangular shape approximately 10 km in length and 1 km in width, was situated about 7 km beneath the southeastern plain and showed a volume change of 68 × 106 m3. The volume increase of the Sill source was approximately twice that of the volume decrease of the Mogi source. Results of this study suggest that magma intrusion into the southeastern plain, originating from the Mogi source and deeper magma reservoirs, likely drove the 2017 unrest. This highlights the complex magmatic activity of Cerro Azul volcano and underscores the importance of continuous displacement monitoring to assess volcanic activity and associated hazards.

Time-series Normalized Difference Vegetation Index (NDVI) data extracted from optical satellite images is the most widely used data for predicting crop growth and yield. In the time series pattern of NDVI, variation due to the growth cycle of crops and reduction due to clouds are often seen. In particular, the rainy season in Korea is an important growing period for crops, but it is very difficult to use optical satellite images due to many clouds. In this study, a spatio-temporal gap-filling technique was developed for Sentinel-2A/B NDVI images obtained in rice paddy in Dangjin. The spatial gap-filling technique is used to correct the boundary of a missing area or a small-area missing area by obtaining the NDVI information of normal pixels at the periphery. Afterward, pixels in large areas of missing areas are temporally gap-filled by applying a Gaussian Process Regression (GPR) model to data acquired before and after the target period. The spatio-temporal gap-filling technique developed in this study showed a Root Mean Squared Error (RMSE) of less than 0.15 for the periodically composited NDVI image, and the clouds were removed and the corrected image could be confirmed with visual interpretation. In addition, for daily NDVI images with a large number of clouds, it was found that RMSE was 0.11 to 0.17 depending on the amount of clouds. The algorithm was developed as a program using Python and takes less than 5 minutes to process. In the future, it will be provided to experts who use agricultural and forestry satellite images. This algorithm will be tested on a variety of land cover areas, including mountainous areas as well as agricultural areas, and will be applied to longer time series data. Through this, we plan to analyze the applicability not only to the vegetation index but also to other biophysical variables required for crop monitoring.

Remote sensing technology has significantly enhanced crop growth and disease monitoring. While satellites operate above the clouds, Unmanned Aerial Vehicle (UAV) mostly operate below the clouds, making cloud cover a critical factor. This study aims to identify the best surface reflectance correction method for UAV images under Sky Clear (SKC) and Broken Clouds Sky (BKN) conditions by comparing NoProcess, New Row Gradient (NewROW), and Bidirectional Reflectance Distribution Function (BRDF) methods. Data was collected using a RedEdge MX camera, and the accuracy of the data results was validated by comparing them with Calibrated Reference Tarp (CRT) ground reflectance data. The RedEdge MX camera includes the Original Row Gradient (OriROW) parameter in its metadata. For users without the BRDF parameter (f_iso, f_geo, and f_vol), OriROW can be utilized after data collection. NewROW was proposed to address the limitations of existing OriROW. The NewROW formula was enhanced to consider sun position and image center values. In the SKC, NewROW and BRDF were highly accurate with correction total values of ±1.57 and ±1.68% in all bands, while NoProcess was lower accurate with correction values of ±4.02%. In BKN, NewROW, and BRDF performed well with correction values of 1.31 and 1.33%, while NoProcess was less effective at 1.83%. Comparing the NoProcess, NewROW, and BRDF under SKC and BKN, the reflectance differences were –2.414 to 1.212, –0.687 to 0.745, and –0.989 to 1.143%. Therefore, both BRDF and NewROW showed high accuracy under both clear and cloudy sky conditions, with the simpler NewROW serving as an effective alternative to BRDF.

Barings designs, secures, operates, and licenses satellites to monitor its members' environments and produces large amounts and types of satellite data, all publicly available. Ownership, usage rights, and interest in derivative works of information acquired with significant national legal input also present different competition across countries. The purpose of this study is to analyze the current status of satellite image use, licensing policy, and pricing policy and to clarify topics that require discussion. We investigated overseas cases, including advanced countries that distribute satellite images, investigated and analyzed the level of related laws and regulations, and whether it was free of charge or not, and also included cases of video distribution and sharing in Korea. The opinions of 20 satellite experts were collected and organized to comprehensively organize the issues. Satellite image information such as public information, restrictive requirements, recovery policies such as fee collection for the added value of processed data, expectations and concerns about the revitalization of the industry, etc. were confirmed. Consistent guidance on price and licensing policy is needed, and the space industry and information changes in technology must be taken into account.

The increasing uncertainty in crop production caused by climate change underscores the necessity of accurate yield predictions for staple crops like rice. This study developed a rice yield prediction model that accounts for climate change by integrating satellite imagery and artificial intelligence. Key variables, including temperature, precipitation, and solar radiation, were derived from Korea Meteorological Administration Automated Synoptic Observing System (ASOS) data and combined with NASA’s Shuttle Radar Topography Mission Digital Elevation Model (SRTM DEM) to incorporate topographic factors. A comprehensive database integrating vegetation indices from Moderate Resolution Imaging Spectroradiometer (MODIS) and Visible Infrared Imaging Radiometer Suite (VIIRS) was established, and Automated Machine Learning (AutoML) was employed to optimize the rice yield prediction model. Validation results demonstrated high predictive accuracy, with the model utilizing ASOS data achieving a MAPE of 4.091% in 2023. This study contributes to data-driven decision-making for climate-resilient agricultural policies and enhanced food security.

Deep learning technologies have revolutionized image processing and analysis, introducing groundbreaking innovations that significantly improve the accuracy and efficiency of object segmentation, especially in satellite imagery. The increasing availability of high-resolution satellite images has created a demand for advanced models capable of handling the complexities of object detection in diverse environments. This study investigates the potential of the Segment Anything Model for Remote Sensing (SAMRS), a deep learning framework specifically designed for remote sensing applications, to accurately identify and segment a wide range of objects within satellite imagery. The model was trained using prominent datasets such as Dataset for Object Detection in Aerial Images (DOTA), Dataset for Object Detection in Optical Remote Sensing Images (DIOR), Fine-grained Object Detection in Aerial Images for Remote Sensing Version 2.0 (FAIR1M-2.0), and Instance Segmentation in Aerial Images Dataset (iSAID), enabling it to learn diverse object features and complexities. The evaluation of SAMRS was conducted on Northwestern Polytechnical University Very High Resolution 10-Class Dataset (NWPU VHR-10) and Beijing-3B datasets, where it demonstrated impressive results. In vehicle detection tasks, SAMRS achieved an Intersection over Union (IoU) of 0.9175, an F1-score of 0.9570, and an accuracy of 0.9385. These metrics highlight SAMRS’s capability to automate object detection in complex satellite images, overcoming challenges posed by intricate backgrounds and diverse object sizes. Furthermore, SAMRS is optimized to analyze both large and small-scale objects, ensuring robust performance across varying conditions. The findings emphasize the model’s utility not only for current remote sensing applications but also for future extensions involving drone imagery and domestic satellite datasets. By automating object detection and segmentation, SAMRS has the potential to transform practical fields such as urban planning, disaster management, traffic monitoring, and environmental analysis, making it a vital tool in advancing satellite imagery analysis.

As the use of drone-based hyperspectral data increases, atmospheric correction methods that are suitable for the characteristics of drone data are required. In this letter, we evaluated the performance of the drone atmospheric correction (DROACOR) model for atmospheric correction of drone hyperspectral data by comparing it with the results of empirical line correction. The results using the normalized difference vegetation index and Fe3+ index showed a high correlation (R2=0.99) between the two correction models in the visible/near-infrared (VNIR). However, the DROACOR model estimated the overall reflectance of kaolin in the short-wave infrared (SWIR) relatively low and the characteristic absorption depth occurring around 2200 nm weakly. These results show the high performance of the DROACOR model in VNIR and its limitations in SWIR.

This study evaluated the performance of LiDAR sensors by mounting the DJI Zenmuse L1 and L2 on a DJI Matrice 350 RTK drone to capture ground point cloud data over flat terrain and forested areas. Scans were conducted at altitudes of 50 m and 80 m, with 70 cm × 70 cm aerial targets placed on open ground and under forest vegetation to measure point density. We assessed the multiple reflections in the captured LiDAR data and reviewed absolute accuracy using ground control points. Results showed that the L2 sensor achieved higher point density than the L1 sensor in open areas and performed well at low altitudes (50 m). For simple terrain like flat ground, adequate data was obtained with only the first or second reflections. In forested areas, the L2 sensor effectively captured ground data beneath dense vegetation, whereas the L1 sensor struggled to accurately detect ground under these conditions. In terms of absolute accuracy, the L2 sensor showed less noise and higher precision on flat, stable terrain than the L1. On more complex terrain, the L2 enabled accurate 3D modeling at higher point density. Overall, the flexible use of both L1 and L2 sensors is expected to improve disaster site analysis and response across varied landscapes.

Rural facilities such as factories and barns near residential areas cause pollution and civil complaints. They are managed through legal registration, but there are still conflicts and inconsistencies. In Korea, the Rural Spatial Restructuring Act will come into effect in March this year. It will promote long-term rural area construction plans and support new public businesses. In this situation, it is essential to have a means to accurately and efficiently identify the status and changes in rural facilities. In this study, we constructed a dataset containing the Korea Multi-purpose Satellite (KOMPSAT) optical images and associated rural facility masks for four cities in South Korea in 2019 and 2020. A deep learning-based segmentation method was then applied to the dataset. Satellite images only show roofs, making it inherently difficult to completely distinguish between building types. Non-target buildings often look almost identical to the target. The higher the complexity of the deep neural network architecture, the more likely these inconsistencies are to cause overfitting problems. For better universal performance, we constructed redundant models from different combinations of data. Redundant models produce different inference results for the same validation sample. Averaging this gives more reliable results. Finally, we also performed a performance comparison between the original model and the new model optimized within the trust region guaranteed by the redundant models.

Advances in the aerospace industry have driven growing research into the use of artificial intelligence for analyzing objects of interest in satellite imagery. Unlike typical 8-bit RGB camera images, however, satellite imagery often contains 16-bit pixel values, which can result in outliers that darken the image. This issue leads to difficulty in object identification and negatively impacts analysis performance. To address this issue, various image enhancement techniques have been proposed, but the effectiveness of each technique depends on the specifics of each satellite and the task to be performed. To address this issue, various image enhancement techniques have been proposed. However, since each satellite has unique characteristics and the effectiveness of each technique varies depending on the specific task, it is necessary to carefully evaluate which technique is most suitable. This research analyzed which of the five image enhancement techniques is most suitable for semantic segmentation tasks using the dataset from the KOMPSAT-3A satellite, which is widely used in South Korea. Experimental results using five semantic segmentation models indicated that percentile stretching performed well in three models, suggesting it as the most universally applicable method. In addition, for buildings and roads, which are important objects in urban analysis, recursive separated and weighted histogram equalization (RSWHE) and percentile stretching were found to be effective.

현대 사회는 갈수록 대형화되는 자연재해와 잦은 재난사고에 의한 인적·사회적 피해가 해마다 증가하고 있다. 난접근 지역이거나 접근 불능의 위험한 재난 현장을 인공위성이나 드론, 조사로봇과 같은 첨단 조사장비를 활용하여 신속하게 접근하고 유의미한 재난 정보를 적시적으로 수집·분석함으로써, 사전 예방·대비 대책 마련뿐 만 아니라 적절한 재난 현장 대응 및 중장기적 복구 계획 수립 등 재난관리 전주기에 걸쳐 국민의 재산과 생명을 지킬 수 있는 중차대한 역할을 수행할 수 있다. 본 특별호에서는 지구 원격 관측 수단인 인공위성 기술뿐만 아니 라 근거리 재난현장 관측센서가 탑재된 이동형 조사차량, 드론, 조사로봇 등 다양한 조사 플랫폼을 활용한 연구원 의 재난관리 현업화 기술을 소개하고 있다. 주요 연구 성과로 구글어스 엔진을 활용한 수재해 피해 탐지와 중·장 기적 시계열 관측, Sentinel-1 Synthetic Aperture Radar (SAR) 영상과 인공지능을 활용한 저수지 수체 탐지, 산불 재 난시 주민 이동 패턴 분석과 재난안전 연구 데이터의 효율적인 통합 관리와 활용방안 연구성과를 소개하였다. 아 울러, 접근 불능의 위험한 재난현장 조사시 드론, 조사로봇을 활용한 재난원인 과학조사 연구성과를 기술하였다.