• 제목/요약/키워드: Remote Sensing Information Models

검색결과 211건 처리시간 0.038초

기계학습 기반의 산불위험 중기예보 모델 개발 (Development of Mid-range Forecast Models of Forest Fire Risk Using Machine Learning)

  • 박수민;손보경;임정호;강유진;권춘근;김성용
    • 대한원격탐사학회지
    • /
    • 제38권5_2호
    • /
    • pp.781-791
    • /
    • 2022
  • 산불로 인한 피해를 최소화하기 위해서 산불위험 예보 정보를 제공하는 것은 필수적이다. 따라서 본 연구에서는 우리나라를 대상으로 기계학습 기반의 산불위험 중기예보(1일 후부터 7일 후까지) 모델을 개발하였다. Global Data Assimilation and Prediction System (GDAPS)의 기상예보 자료와 기 개발된 산불위험지수(Fire Risk Index, FRI)의 과거 및 현재 정보, 그리고 기타 환경요소(i.e., 고도, 산불다발지수, 가뭄지수)의 현재 정보를 반영하여 모델을 개발하였다. 본 연구에서는 실시간 학습을 통해 모델을 개발하였으며, 효율적인 모델 개발을 목적으로 과거 산불위험지수와 가뭄지수의 유무를 고려하여 세가지 경우(Scheme 1: 과거 산불위험지수 및 가뭄지수, Scheme 2: 과거 산불위험지수, Scheme 3: 과거 산불위험지수 변화 추세 및 가뭄지수)로 연구를 수행하였다. 본 연구에서 개발된 산불위험예보모델은 예보기간에 상관없이 높은 정확도(피어슨 상관계수(Pearson correlation) >0.8, relative root mean square error <10%)를 나타냈으며, 실제 산불 발생 건에 대해서도 유의미한 결과를 보였다. 과거 산불위험지수의 추세보다는 산불위험지수 값 자체를 입력변수로 사용하는 것이 높은 정확도를 보였으며, 가뭄지수 사용과 관계없이 좋은 결과를 나타냈다.

Biomass Estimation of Gwangneung Catchment Area with Landsat ETM+ Image

  • Chun, Jung Hwa;Lim, Jong-Hwan;Lee, Don Koo
    • 한국산림과학회지
    • /
    • 제96권5호
    • /
    • pp.591-601
    • /
    • 2007
  • Spatial information on forest biomass is an important factor to evaluate the capability of forest as a carbon sequestrator and is a core independent variable required to drive models which describe ecological processes such as carbon budget, hydrological budget, and energy flow. The objective of this study is to understand the relationship between satellite image and field data, and to quantitatively estimate and map the spatial distribution of forest biomass. Landsat Enhanced Thematic Mapper (ETM+) derived vegetation indices and field survey data were applied to estimate the biomass distribution of mountainous forest located in Gwangneung Experimental Forest (230 ha). Field survey data collected from the ground plots were used as the dependent variable, forest biomass, while satellite image reflectance data (Band 1~5 and Band 7), Normalized Difference Vegetation Index (NDVI), Soil-Adjusted Vegetation Index (SAVI), and RVI (Ratio Vegetation Index) were used as the independent variables. The mean and total biomass of Gwangneung catchment area were estimated to be about 229.5 ton/ha and $52.8{\times}10^3$ tons respectively. Regression analysis revealed significant relationships between the measured biomass and Landsat derived variables in both of deciduous forest ($R^2=0.76$, P < 0.05) and coniferous forest ($R^2=0.75$, P < 0.05). However, there still exist many uncertainties in the estimation of forest ecosystem parameters based on vegetation remote sensing. Developing remote sensing techniques with adequate filed survey data over a long period are expected to increase the estimation accuracy of spatial information of the forest ecosystem.

연안 해저 재질 분석을 위한 초분광영상의 보정 방법 (Method of Correcting Hyperspectral Image for Seabed Material Analysis of Coastal Area)

  • 신명식;신정일;김익재;서용철
    • 한국지리정보학회지
    • /
    • 제19권2호
    • /
    • pp.107-116
    • /
    • 2016
  • 연안 해저 재질 조사에 있어 위성 및 항공 원격탐사 자료를 이용하면 기존 현장조사 방법에 비해 효율성을 높일 수 있으나, 물에 의한 빛의 흡수 특성으로 인해 동일한 조건이더라도 수심에 따라 영상에서 다른 반사도를 보이게 된다. 따라서 본 연구에서는 초분광영상을 연안 해저 재질 분석 자료로 사용하기 위한 보정 방법을 제시하고자 한다. 연구지역은 강원도 사천진항에서 경포해수욕장 일대이고, 사용한 초분광영상은 CASI-1500 영상이다. 수체 반사율과 수심 간의 회귀모델을 통해 밴드별 산란흡수계수를 추정하여 영상에 적용하였다. 그 결과 수심보정 전 영상에서 매우 얕은 수심에 한정하여 판독이 가능하였지만 수심보정 후 상대적으로 깊은 수심까지 판독이 가능해지고, 수심에 따른 해저면의 반사율 변이가 크게 감소한 것을 알 수 있었다.

기름이 유출된 바다 표면의 L-밴드 전파 산란에 대한 수치해석적 연구 (Numerical Simulation of Radar Backscattering from Oil Spills on Sea Surface for L-band SAR)

  • 박성민;양찬수;오이석
    • 대한원격탐사학회지
    • /
    • 제26권1호
    • /
    • pp.21-27
    • /
    • 2010
  • 본 논문에서는 기름이 유출된 바다 표면의 레이더 산란에 대한 수치해석적 연구를 보여준다. 우선, 풍속에 따라서 불규칙적인 거칠기를 갖는 바다 표면을 생성한 다음에, 유전율이 높은 거친 바닷물 표면 위에 유전율이 낮은 기름층을 두어 기름이 유출된 바다 표면을 생성한다. 서로 다른 표면 거칠기, 기름층의 두께와 유전율, 주파수, 편파, 입사각의 조합으로 이루어진 다양한 형태의 기름 유출 바다에 대한 레이더 후방산란계수를 모멘트 법(Method of Moments)/ 몬테카를로(Monte-Carlo) 방법을 이용하여 계산한다. 이 수치해석적 방법은 이론적인 산란 모델로 계산 가능한 간단한 구조에 대해서 이론 모델 결과와 비교함으로써 그 정확성을 검증한다. 이 수치해석적 방법으로 기름이 유출된 바다 표면에서의 후방산란계수의 감소를 분석하며, 이 분석 결과는 결과적으로 SAR 영상에서의 기름층의 발견 및 식별에 도움을 줄 것이다.

지표면 별 영상잡음과 영상질감을 이용한 SAR 클러터 영상 생성 (SAR Clutter Image Generation Based on Measured Speckles and Textures)

  • 권순구;오이석
    • 대한원격탐사학회지
    • /
    • 제25권4호
    • /
    • pp.375-381
    • /
    • 2009
  • 본 논문에서는 다양한 종류의 지표면에 대하여 분석하여 산란 특성을 연구하고 SAR 클러터 영상을 제작하고 실제 SAR 클러터 영상과 비교한다. 먼저 지표면의 특성을 분석하기 위해 각각의 지표면에 대해서 입력변수를 측정한다. 측정한 데이터를 이용하여 Oh 모델, PO 모델, radiative transfer model(RTM)을 이용하여 각도 별 산란계수를 구하였다. SAR 영상 생성을 위해 먼저 측정 지역의 DEM (digital elevation map)과 LCM (land cover map)데이터를 제작한다. DEM 데이터의 단일 픽셀(pixel)의 높이 정보를 이용하여 픽셀의 입사각을 계산하고 입사각에 따른 해당 지표면의 산란 계수를 대입한다. LCM 데이터는 해당 지역의 답사를 통해 논, 밭, 산, 길, 인공물 등을 1:5000 지도에 기입하여 SAR 영상 생성에 사용한다. DEM 데이터와 LCM 데이터를 사용하여 입사각과 지표면 종류에 따른 계수를 계산하고 영상잡음(speckle)과 영상질감(texture)을 이용하여 SAR 클러터 영상을 생성하고 실제 영상과 비교한다.

퍼지집합과 베이지안 확률 기법을 이용한 암설사면지형 분포지역 추출에 관한 연구 (The Study on the Extraction of the Distribution Potential Area of Debris Landform Using Fuzzy Set and Bayesian Predictive Discriminate Model)

  • 위눈솔;장동호
    • 한국지형학회지
    • /
    • 제24권3호
    • /
    • pp.105-118
    • /
    • 2017
  • The debris slope landforms which are existent in Korean mountains is generally on the steep slopes and mostly covered by vegetation, it is difficult to investigate the landform. Therefore a scientific method is required to come up with an effective field investigation plan. For this purpose, the use of Remote Sensing and GIS technologies for a spatial analysis is essential. This study has extracted the potential area of debrisslope landform formation using Fuzzy set and Bayesian Predictive Discriminate Model as mathematical data integration methods. The first step was to obtain information about debris locations and their related factors. This information was verified through field investigation and then used to build a database. In the second step, the map that zoning the study area based on the degree of debris formation possibility was generated using two modeling methods, and then cross validation technique was applied. In order to quantitatively analyze the accuracy of two modeling methods, the calculated potential rate of debrisformation within the study area was evaluated by plotting SRC(Success Rate Curve) and calculating AUC(Area Under the Curve). As a result, the prediction accuracy of Fuzzy set model wes 83.1% and Bayesian Predictive Discriminate Model wes 84.9%. It showed that two models are accurate and reliable and can contribute to efficient field investigation and debris landform management.

Cloud Removal Using Gaussian Process Regression for Optical Image Reconstruction

  • Park, Soyeon;Park, No-Wook
    • 대한원격탐사학회지
    • /
    • 제38권4호
    • /
    • pp.327-341
    • /
    • 2022
  • Cloud removal is often required to construct time-series sets of optical images for environmental monitoring. In regression-based cloud removal, the selection of an appropriate regression model and the impact analysis of the input images significantly affect the prediction performance. This study evaluates the potential of Gaussian process (GP) regression for cloud removal and also analyzes the effects of cloud-free optical images and spectral bands on prediction performance. Unlike other machine learning-based regression models, GP regression provides uncertainty information and automatically optimizes hyperparameters. An experiment using Sentinel-2 multi-spectral images was conducted for cloud removal in the two agricultural regions. The prediction performance of GP regression was compared with that of random forest (RF) regression. Various combinations of input images and multi-spectral bands were considered for quantitative evaluations. The experimental results showed that using multi-temporal images with multi-spectral bands as inputs achieved the best prediction accuracy. Highly correlated adjacent multi-spectral bands and temporally correlated multi-temporal images resulted in an improved prediction accuracy. The prediction performance of GP regression was significantly improved in predicting the near-infrared band compared to that of RF regression. Estimating the distribution function of input data in GP regression could reflect the variations in the considered spectral band with a broader range. In particular, GP regression was superior to RF regression for reproducing structural patterns at both sites in terms of structural similarity. In addition, uncertainty information provided by GP regression showed a reasonable similarity to prediction errors for some sub-areas, indicating that uncertainty estimates may be used to measure the prediction result quality. These findings suggest that GP regression could be beneficial for cloud removal and optical image reconstruction. In addition, the impact analysis results of the input images provide guidelines for selecting optimal images for regression-based cloud removal.

Use of Fuzzy Object Concept in GIS-based Spatial Prediction Model for Landslide Hazard Mapping

  • Park, No-Wook;Chi, Kwang-Hoon
    • 대한원격탐사학회:학술대회논문집
    • /
    • 대한원격탐사학회 2002년도 Proceedings of International Symposium on Remote Sensing
    • /
    • pp.123-127
    • /
    • 2002
  • In this paper, we propose spatial prediction model for landslide hazard mapping that can account for the fuzziness of boundaries in thematic maps showing the different environmental impacts, depending on the scales and the resolutions of them. The fuzziness or uncertainty of boundary is represented in favourability function based on fuzzy object concept and the effects of them are quantitatively evaluated with the help of cross validation procedures. To illustrate the proposed schemes, a case study from Boeun, Korea was carried out. As a result, the proposed schemes are helpful to account for intrinsic uncertainties in categorical maps and can be effectively adopted in spatial prediction models for other purposes.

  • PDF

A STUDY ON DEM GENE]RATON USING POLYNOMIAL CAMERA MODEL IN SATELLITE IMAGERY

  • Jeon, Seung-Hun;Kim, Sung-Chai;Lee, Heung-Jae;Lee, Kae-hei
    • 대한원격탐사학회:학술대회논문집
    • /
    • 대한원격탐사학회 2002년도 Proceedings of International Symposium on Remote Sensing
    • /
    • pp.518-523
    • /
    • 2002
  • Nowadays the Rational Function Model (RFM), an abstract sensor model, is substituting physical sensor models for highly complicated imaging geometry. But RFM is algorithm to be required many Ground Control Points (GCP). In case of RFM of the third order, At least forty GCP are required far RFM generation. The purpose of this study is to research more efficient algorithm on GCP and accurate algorithm similar to RFM. The Polynomial Camera Model is relatively accurate and requires a little GCP in comparisons of RFM. This paper introduces how to generate Polynomial Camera Model and fundamental algorithms for construction of 3-D topographic data using the Polynomial Camera Model information in the Kompsat stereo pair and describes how to generate the 3-D ground coordinates by manual matching. Finally we tried to extract height information for the whole image area with the stereo matching technique based on the correlation.

  • PDF

EXTRACTING COMPLEX BUILDING FROM AIRBORNE LIDAR AND AIRBORNE ORTHIMAGERY

  • Nguyen, Dinh-Tai;Lee, Seung-Ho;Cho, Hyun-Kook
    • 대한원격탐사학회:학술대회논문집
    • /
    • 대한원격탐사학회 2008년도 International Symposium on Remote Sensing
    • /
    • pp.177-180
    • /
    • 2008
  • Many researches have been tried to extract building models and created a 3D cyber city from LiDAR data. In this paper, the approach of extracting complex building by using airborne LiDAR data combined with airborne orthoimagery has been performed. The pseudo-building elevations were derived from modified discrete return LiDAR data. Based on information property of the pseudo-height, building features could be extracted. The results of this study indicated the improvement of building extraction.

  • PDF