Evaluation of NDVI Estimation Considering Atmospheric and BRDF Correction through Himawari-8/AHI

성노훈,정대성,김진수,한경수 한국기상학회 2020 Asia-Pacific Journal of Atmospheric Sciences Vol.56 No.2

Satellite-based vegetation indices are an essential element in understanding the Earth’s surface. In this study, we estimated the normalized difference vegetation index (NDVI) using Himawari-8/Advanced Himawari Imager (AHI) data and analyzed the sensitivity of products to atmospheric and surface correction. We used the Second Simulation of a Satellite Signal in the Solar Spectrum (6S) radiative transfer model for atmospheric correction, and kernel-based semi-empirical bidirectional reflectance distribution function (BRDF) model to remove surface anisotropic effects. From this, top-of-atmosphere, top-of-canopy, and normalized NDVIs were produced. A sensitivity analysis showed that the normalized NDVI had the lowest number of missing values compared with the others and almost no low peaks during the study period. These results were validated by Terra and Aqua/Moderate Resolution Imaging Spectroradiometer (MODIS) and Project for On-Board Autonomy/Vegetation (PROBA) NDVI product, showing the root mean square error (RMSE) and bias of 0.09 and + 0.04 (MODIS) and 0.09 and − 0.04 (PROBA), respectively. These results also satisfied the FP7 Geoland2/BioPar project-defined user requirements (threshold: 0.15; target: 0.10).

LTE-A 이종망 환경에서 ABS와 Power Control의 동시적용과 ABS 비율 도출

성노훈(No-Hoon Sung),최영준(Young-June Choi),장지웅(Ji-Woong Jang) 한국통신학회 2015 韓國通信學會論文誌 Vol.40 No.12

NDWI를 활용한 한반도 지역의 산림 캐노피에 대한 water stress 평가

성노훈 ( No Hun Seong ),서민지 ( Min Ji Seo ),이경상 ( Kyeong Sang Lee ),이창석 ( Chang Suk Lee ),김현지 ( Hyun Ji Kim ),최성원 ( Sung Won Choi ),한경수 ( Kyung Soo Han ) 대한원격탐사학회 2015 大韓遠隔探査學會誌 Vol.31 No.2

잎의 수분 함유량은 식물의 건강상태를 나타내는 중요한 척도 중 하나로써, 이를 원격탐사를 활용하여 모니터링 하는 것은 산림관리에 있어서 매우 중요하다. 본 연구에서는 식생 캐노피의 수분량을 연구하는데 유용한 지수인 Normalized Difference Water Index (NDWI)를 이용하여 한반도 산림의 water stress 정도를 알아보고자 한다. SPOT/VEGETATION S10 채널자료를 1999년부터 2013년까지 취득하여 NDWI 를 산출하였고, 데이터의 노이즈를 제거하기 위하여 단순이동평균, NDWI의 시간적 변화를 파악하기 위하여standardized anomaly를 수행했으며, 직관적인 모니터링을 위해 NDWI anomaly를 등급화 하였다. 또한 피해면적 150 ha 이상의 대형 산불과 비교 검증을 통해, 산림 캐노피의 water stress 평가 인자로서 Leaf water content is one of important indicators that shows states of vegetation. It is important to monitor vegetation water content using remote sensing for forest management. In this study, we investigated the degree of water stress in Korean peninsula with Normalized Difference Water Index (NDWI) to study the water content of vegetation canopy. We calculated the NDWI using SPOT/ VEGETATION S10 channel data over forest from 1999 to 2013. We calculated Simple Moving Average (SMA) to remove temporal noises of NDWI in time series, and used standardized anomaly to investigate temporal changes. We classified the NDWI anomalies into three scales (low, moderate, and high) in order to monitor intuitively. We also investigated suitability of the NDWI as an evaluation criterion about water stress of vegetation canopy by comparing and verifying forest fires damaged area over 150 ha. Consequently, huge forest fire occurred 24 times during the study period. Also, negative anomalies appeared in every forest fire location and their neighboring areas. In particular, we found huge forest fires where NDWI anomalies were in ‘high’ scale.

LTE-A 이종망 환경에서 ABS와 Power Control의 동시적용과 효율적인 ABS 비율 도출

성노훈(Sung No-Hoon),최영준(Choi Young-June) 한국통신학회 2015 한국통신학회 학술대회논문집 Vol.2015 No.06

Landsat-8을 활용한 Sentinel-2A Near Infrared 채널의 Spectral Band Adjustment Factor 적용성 평가

김나연,성노훈,정대성,심수영,우종호,최성원,박성우,한경수 대한원격탐사학회 2023 大韓遠隔探査學會誌 Vol.39 No.3

Various earth observation satellites need to provide accurate and high-quality data afterlaunch. To maintain and enhance the quality of satellite data, it is crucial to employ a cross-calibrationprocessthat accountsfor differencesin sensor characteristics,such asthe spectral band adjustment factor(SBAF). In this study, we utilized Landsat-8 and Sentinel-2A satellite imagery collected from desertsites in Libya4, Algeria3, and Mauritania2 among pseudo-invariant calibration sites to calculate andapply SBAF, thereby compensating the uncertainties arising from variations in bandwidths. Wequantitatively compared the reflectance differences based on the similarity of bandwidths, includingBlue, Green, Red, and both the near-infrared (NIR) narrow, and NIR bands of Sentinel-2A. Followingthe application of SBAF, significant results with reflectance differences of approximately 1% or lesswere observed for all bands except NIR. In the case of the Sentinel-2A NIR band, it exhibited asignificantly larger bandwidth difference compared to the NIR narrow band. However, after applyingSBAF, the reflectance difference fell within the acceptable error range (5%) of 1–2%. It indicates that SBAF can be applied even when there is a substantial difference in the bandwidths of the two sensors,particularly in situations where satellite utilization is limited. Therefore, it was determined that SBAFcould be applied even when the bandwidth difference between the two sensors is large in a situationwhere satellite utilization is limited. It is expected to be helpful in research utilizing the quality andcontinuity of satellite data.

북극 해빙표면온도 산출을 위한 Automated Machine Learning과 Deep Neural Network의 적용성 평가

박성우,성노훈,심수영,정대성,우종호,김나연,김홍희,한경수 대한원격탐사학회 2023 大韓遠隔探査學會誌 Vol.39 No.6

This study utilized automated machine learning (AutoML) to calculate Arctic ice surfacetemperature (IST). AutoML-derived IST exhibited a strong correlation coefficient (R) of 0.97 and aroot mean squared error (RMSE) of 2.51K. Comparative analysis with deep neural network (DNN)models revealed that AutoML IST demonstrated good accuracy, particularly when compared toModerate Resolution Imaging Spectroradiometer (MODIS) IST and ice mass balance (IMB) buoy IST. These findings underscore the effectiveness of AutoML in enhancing IST estimation accuracy underchallenging polar conditions.

고해상도 위성자료를 활용한 마른 잎 탐지

심수영,진동현,성노훈,이경상,서민지,최성원,정대성,한경수 대한원격탐사학회 2020 大韓遠隔探査學會誌 Vol.36 No.3

Recently, many studies have been conducted on the changing phenology on the Korean Peninsula due to global warming. However, because of the geographical characteristics, research on plant season in autumn, which is difficult to measure compared to spring season, is insufficient. In this study, all leaves that maple and fallen leaves were defined as ‘Decay leaves’ and decay leaf detection was performed based on the Landsat-8 satellite image. The first threshold value of decay leaves was calculated by using NDVI and the secondary threshold value of decay leaves was calculated using by NDWI and the difference of spectral characteristics with green leaves. POD, FAR values were used to verify accuracy of the dry leaf detection algorithm in this study, and the results showed high accuracy with POD of 98.619 and FAR of 1.203. 최근 지구 온난화의 영향으로 변화하는 한반도 식물계절에 대한 연구가 많이 이루어지고 있다. 그러나지리적인 특성상 봄철 식물계절에 비해 실측이 어려운 가을철 식물계절의 연구는 미비한 실정이다. 이에 본 연구에서는 대표적인 가을철 식물계절인 단풍과 낙엽 등을 ‘마른 잎’으로 정의하고 Landsat-8 위성영상을 기반으로 마른 잎 탐지를 수행하였다. NDVI를 이용하여 마른 잎의 1차 경계 값을 산출하고, 건강한 잎과의 분광특성차이 및 NDWI를 이용하여 마른 잎의 2차 경계 값을 산출하였다. 본 연구의 마른 잎 탐지 알고리즘의 정확도 검증을 위해 POD, FAR 값을 이용하였으며, 검증 결과 POD는 98.619, FAR은 1.203으로 높은 정확성을 보였다.

기계학습을 통한 주간 반투명 구름탐지 연구: GK-2A/AMI를 이용하여

변유경,진동현,성노훈,우종호,전우진,한경수,Byeon, Yugyeong,Jin, Donghyun,Seong, Noh-hun,Woo, Jongho,Jeon, Uujin,Han, Kyung-Soo 대한원격탐사학회 2022 大韓遠隔探査學會誌 Vol.38 No.6

Clouds are composed of tiny water droplets, ice crystals, or mixtures suspended in the atmosphere and cover about two-thirds of the Earth's surface. Cloud detection in satellite images is a very difficult task to separate clouds and non-cloud areas because of similar reflectance characteristics to some other ground objects or the ground surface. In contrast to thick clouds, which have distinct characteristics, thin transparent clouds have weak contrast between clouds and background in satellite images and appear mixed with the ground surface. In order to overcome the limitations of transparent clouds in cloud detection, this study conducted cloud detection focusing on transparent clouds using machine learning techniques (Random Forest [RF], Convolutional Neural Networks [CNN]). As reference data, Cloud Mask and Cirrus Mask were used in MOD35 data provided by MOderate Resolution Imaging Spectroradiometer (MODIS), and the pixel ratio of training data was configured to be about 1:1:1 for clouds, transparent clouds, and clear sky for model training considering transparent cloud pixels. As a result of the qualitative comparison of the study, bothRF and CNN successfully detected various types of clouds, including transparent clouds, and in the case of RF+CNN, which mixed the results of the RF model and the CNN model, the cloud detection was well performed, and was confirmed that the limitations of the model were improved. As a quantitative result of the study, the overall accuracy (OA) value of RF was 92%, CNN showed 94.11%, and RF+CNN showed 94.29% accuracy.

Article KOMPSAT-3/3A 채널별 6SV 조견표의 지표반사도 민감도 분석

정대성,진동현,성노훈,이경상,서민지,최성원,심수영,한경수,김보람 대한원격탐사학회 2020 大韓遠隔探査學會誌 Vol.36 No.5

The radiance measured from satellite has noise due to atmospheric effect. Atmospheric correction is the process of calculating surface reflectance by removing atmospheric effect and surface reflectance is calculated by the Radiative Transfer Model (RTM)-based Look-Up Table (LUT). In general, studies using a LUT make LUT for each channel with the same atmospheric and geometric conditions. However, atmospheric effect of atmospheric factors do not react sensitively in the same channel. In this study, the LUT for each channel of Korea Multi-Purpose SATellite (KOMPSAT)-3/3A was made under the same atmospheric·geometric conditions. And, the accuracy of the LUT was verified by using the simulated Top of Atmosphere radiation and surface reflectance in the RTM. As a result, the relative error of the surface reflectance in the blue channel that sensitive to the aerosol optical depth was 81.14% at the maximum, and 42.67% in the NIR (Near Infrared) channel. 대기효과로 인해 위성에서 측정된 복사휘도는 오차를 가지고 있다. 대기보정은 대기효과를 제거하여지표반사도를 산출하는 과정이며, 지표반사도는 복사전달모델 기반의 조견표(Look-Up Table; LUT)를 통해 산출된다. 일반적으로 조견표를 사용하는 연구들은 동일한 대기·기하조건으로 채널별 조견표를 구축하고 있다. 하지만, 대기 조건들이 민감하게 반응하는 채널은 모두 다르다. 이에 본 연구에서는 동일한 대기·기하조건으로 KOMPSAT-3/3A의 채널별 조견표를 구축하고, 복사전달모델에서 모의된 대기상단 복사휘도 및 지표반사도를 검증 자료로 활용하여 조견표의 정확도를 확인하였다. 결과적으로, 에어로졸 광학 두께에 민감하게 반응하는 Blue 채널에서 지표반사도의 상대오차가 최대 81.14%으로 나타났고, NIR 채널에서는 최대 42.67%으로나타났다.

A Ship-Wake Joint Detection Using Sentinel-2 Imagery

전우진,진동현,성노훈,정대성,심수영,우종호,변유경,김나연,한경수 대한원격탐사학회 2023 大韓遠隔探査學會誌 Vol.39 No.1

Ship detection is widely used in areas such as maritime security, maritime traffic, fisheriesmanagement, illegal fishing, and border control, and ship detection is important for rapid response and damageminimization as ship accident rates increase due to recent increases in international maritime traffic. Currently,according to a number of global and national regulations, ships must be equipped with automatic identificationsystem (AIS), which provide information such as the location and speed of the ship periodically at regularintervals. However, most small vessels (less than 300 tons) are not obligated to install the transponder and maynot be transmitted intentionally or accidentally. There is even a case of misuse of the ship’s location information. Therefore, in this study, ship detection was performed using high-resolution optical satellite images that canperiodically remotely detect a wide range and detect small ships. However, optical images can cause false-alarmdue to noise on the surface of the sea, such as waves, or factors indicating ship-like brightness, such as cloudsand wakes. So, it is important to remove these factors to improve the accuracy of ship detection. In this study,false alarm was reduced, and the accuracy of ship detection was improved by removing wake. As a ship detectionmethod, ship detection was performed using machine learning-based random forest (RF), and convolutionalneural network (CNN) techniques that have been widely used in object detection fields recently, and ship detectionresults by the model were compared and analyzed. In addition, in this study, the results of RF and CNN werecombined to improve the phenomenon of ship disconnection and the phenomenon of small detection. The shipdetection results of this study are significant in that they improved the limitations of each model while maintainingaccuracy. In addition, if satellite images with improved spatial resolution are utilized in the future, it is expectedthat ship and wake simultaneous detection with higher accuracy will be performed.