• Korean Journal of Remote Sensing
• /
• v.35 no.2
• /
• pp.279-288
• /
• 2019

We evaluated the land cover classification performance of SegNet, which features semantic segmentation of aerial imagery. We selected four semantic classes, i.e., urban, farmland, forest, and water areas, and created 2,000 datasets using aerial images and land cover maps. The datasets were divided at a 8:2 ratio into training (1,600) and validation datasets (400); we evaluated validation accuracy after tuning the hyperparameters. SegNet performance was optimal at a batch size of five with 100,000 iterations. When 200 test datasets were subjected to semantic segmentation using the trained SegNet model, the accuracies were farmland 87.89%, forest 87.18%, water 83.66%, and urban regions 82.67%; the overall accuracy was 85.48%. Thus, deep learning-based semantic segmentation can be used to classify land cover.

• Proceedings of the KSRS Conference
• /
• 2009.03a
• /
• pp.22-25
• /
• 2009

본 논문은 기 구축된 환경부의 중분류 토지피복도를 올해 KOMPSAT-2 화상으로 갱신하기 위한 예비연구에서 분류결과의 정확도 평가를 중점적으로 나타낸다. KOMPSAT-2 화상기반의 중분류 토지피복도는 객체지향의 분류기법을 이용하였고, 경계선 수정의 경우 반자동 기법에 의해 제작되었다. 계통표집과 임의표집에 의한 Kappa 분석에서 계통 표집의 KHAT값은 0.81, 임의표집의 KHAT값은 0.89를 각각 가지므로 거의 완벽의 일치성을 표시한다. 따라서 전술한 지도화 방법을 통해 제작된 KOMPSAT-2 화상의 중분류 토지피복도는 이용자에게 토지피복정보의 신뢰성도 함께 제공한다.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2002.10a
• /
• pp.75-78
• /
• 2002

• Proceedings of the Korea Water Resources Association Conference
• /
• 2010.05a
• /
• pp.1094-1098
• /
• 2010

미계측 유역에서의 홍수량을 추정할 수 있는 방법은 다양하게 제시되고 있으나, 이에 대한 평가나 조사는 사실상 전무하여 수자원 설계실무에 이용할 수 있는 절차나 방법은 극히 제한되어있다. 현재 주로 이용하고 있는 홍수량 추정절차는 강우를 근거로 한 확률강우량법, SCS방법, 단위도법이 국내의 표준방법으로 이용되고 있다. 또한 수치지도 및 위성영상분석 등과 같은 GIS 자료의 구축이 가능해짐에 따라서, 국내에서는 토양의 종류와 피복 형태 그리고 선행강우조건까지 종합적으로 고려하여 해석하는 유출곡선번호(SCS Runoff Curve Number; CN) 방법이 많이 사용되고 있다. 유출량 해석 시 이용되는 CN은 토지이용도 및 토양도와 같은 지형학적 인자에 지배받게 된다. 그러나 현재 우리나라에서 제공하는 토지이용도 및 토양도는 그 종류가 다양하고, 분류방식이 상이하여 활용 자료에 따라 CN이 달라지므로 유출율의 차이가 발생하게 된다. 국내에서 제공되는 다양한 자료를 이용하여 최적의 CN값을 산정하기 위한 연구가 선행된 바있다. 허기술(1987) 등은 우리나라의 정밀토양도에 의한 토양군 분류에 관한 연구를 진행하였으며 조홍제(1997, 2001)는 LANDSAT 위성영상을 이용하여 유역의 토지피복상태를 분류하고 식생지수를 고려하여 CN을 추정하였고, 김경탁(1998, 2003, 2004)은 개략토양도와 정밀토양도를 이용하여 유출모의 실행한 결과를 비교하여 신뢰도가 높다고 판단되는 정밀토양도를 사용한 CN 추정기법의 사용을 제안한 바 있다. 본 연구에서는 GIS를 이용하여 국내에서 활용 가능한 토양도 및 토지이용도의 종류에 따라 총 9개 Case로 안동댐 유역의 CN을 산정하였다.

• Korean Journal of Remote Sensing
• /
• v.32 no.2
• /
• pp.141-153
• /
• 2016

Reliable assessment of Land Use and Land Cover (LULC) changes greatly improves many practical issues in hydrography, socio-geographical research such as the observation of erosion and accretion, coastal monitoring, ecological effects evaluation. Remote sensing imageries can offer the outstanding capability to monitor nature and extent of land and associated changes over time. Nowadays accurate analysis using remote sensing imageries with high spatio-temporal resolution is required for environmental monitoring. This study develops a methodology of mapping and change detection in LULC by using classified Korea Multi-Purpose Satellite-2 (KOMPSAT-2) multispectral imageries at Jeonbuk and Jeonnam provinces including protected tidal flats located in the west coasts of Korean peninsula from 2008 to 2015. The LULC maps generated from unsupervised classification were analyzed and evaluated by post-classification change detection methods. The LULC assessment in Jeonbuk and Jeonnam areas had not showed significant changes over time although developed area was gradually increased only by 1.97% and 4.34% at both areas respectively. Overall, the results of this study quantify the land cover change patterns through pixel based analysis which demonstrate the potential of multispectral KOMPSAT-2 images to provide effective and economical LULC maps in the coastal zone over time. This LULC information would be of great interest to the environmental and policy mangers for the better coastal management and political decisions.

• 한국지구과학회:학술대회논문집
• /
• 2005.09a
• /
• pp.79-84
• /
• 2005

이 논문에서는 구름과 같은 기상 상태의 제약 없이 자료 획득이 가능한 SAR 자료를 이용하여 토지 피복 특성을 구분하고자 하였다. 기존 단일 주파수, 편광 상태의 자료만을 제공하는 SAR 자료를 이용한 분류에서의 낮은 분류 정확도를 향상시키고자 이 논문에서는 다중 시기 C 밴드 자료이면서 서로 다른 편광 상태의 자료를 제공하는 Radarsat-1(HH)와 ENVISAT(VV) 자료를 분류에 이용하였다. 분류 기법으로 Random Forests를 적용한 결과, 단일 편광 상태의 자료만을 이용하였을 때에 비해서 보다 향상된 분류 정확도를 얻을 수 있었다.

• Journal of Korean Society for Geospatial Information Science
• /
• v.19 no.4
• /
• pp.153-159
• /
• 2011

Vegetative land cover is an important variable many Earth system process, general circulation and carbon exchange model requires vegetative cover as boundary layer necessary to run model. The purpose of this study is to detect and to understand land surface change. To monitor changes of East Asia vegetation, we used NDVI 10-day MVC data derived from SPOT VEGETATION during 12 years from 1999 to 2010. Finally, according to the land cover of classified class, we performed analysis for dynamic zone(positive change zone and negative change zone), static zone in 1999, 2010. Therefore, land covers corresponding to each class have appeared change by 2010. Land cover change was confirmed by analyzing data during 12 years which appeared vegetation change of surrounding the actual desert area to east.

• Journal of Korean Society for Geospatial Information Science
• /
• v.9 no.2 s.18
• /
• pp.37-46
• /
• 2001

Spectral pattern recognition techniques are most used in classification of remotely sensed data. Yet, in any real image, adjacent pixels are related, because imaging sensors acquire significant portions of energy from adjacent pixels. And, with the continued improvement in the spatial resolution of remote sensing systems, another spatial pattern recognition approach is must considered. In this study, we aim to show the potentiality of spatial classification methods through comparing the accuracies of spectral classification methods and those of spectral classification methods. By the comparisons between the two methods, classification accuracies of 6 different spatial classification methods are higher than that of spectral classification method by 2-6% or so. Additionally, we can show it statistically through the classification experiments with different band combinations.

• Korean Journal of Remote Sensing
• /
• v.39 no.3
• /
• pp.269-282
• /
• 2023

The purpose of this study was to construct land cover maps using a deep learning model and to select the optimal deep learning model for land cover classification by adjusting the dataset such as input image size and Stride application. Two types of deep learning models, the U-net model and the DeeplabV3+ model with an Encoder-Decoder network, were utilized. Also, the combination of the two deep learning models, which is an Ensemble model, was used in this study. The dataset utilized RapidEye satellite images as input images and the label images used Raster images based on the six categories of the land use of Intergovernmental Panel on Climate Change as true value. This study focused on the problem of the quality improvement of the dataset to enhance the accuracy of deep learning model and constructed twelve land cover maps using the combination of three deep learning models (U-net, DeeplabV3+, and Ensemble), two input image sizes (64 × 64 pixel and 256 × 256 pixel), and two Stride application rates (50% and 100%). The evaluation of the accuracy of the label images and the deep learning-based land cover maps showed that the U-net and DeeplabV3+ models had high accuracy, with overall accuracy values of approximately 87.9% and 89.8%, and kappa coefficients of over 72%. In addition, applying the Ensemble and Stride to the deep learning models resulted in a maximum increase of approximately 3% in accuracy and an improvement in the issue of boundary inconsistency, which is a problem associated with Semantic Segmentation based deep learning models.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.16 no.1
• /
• pp.75-84
• /
• 1998

To achieve more accurate information from satellite image data, a research on a technique for mixed pixel ex-traction has been produced. The mixed pixels with only two land covers have been experimented. By analyzing canonical vector in canonical correlation classification, the mixed pixels have been classified. The ratio of the two canonical weighted values-the elements of canonical vector have been used as a threshold to discriminate mixed pixels. In case of the classification for the mixed pixels of bridge and water class in TM data before or after the 1st of September, the threshold for the optimal classification of the mixed pixels is 4.0. That is, if the ratio of the two canonical weighted values is less than 4.0, the pixel is a mixed pixel. Also, using the distribution of canonical weighted values, the constitution percentages of land covers within one mixed pixel can be approximately deducted. The accuracy of mixed pixel extraction for experimental area is 90％ and quite acceptable. Conclusively, a technique for mixed pixel extraction by canonical vector analysis is effective.