• Journal of Cadastre & Land InformatiX
• /
• v.49 no.1
• /
• pp.103-112
• /
• 2019

Extraction of region of interest for satellite image classification is one of the important techniques for efficient management of the national land space. However, recent studies on satellite image classification often depend on the information of the selected image in selecting the region of interest. This study propose an effective method of selecting the area of interest using the continuous digital topographic map constructed from high resolution images. The spatial information used in this research is based on the digital topographic map from 2013 to 2017 provided by the National Geographical Information Institute and the 2015 Sejong City land cover map provided by the Ministry of Environment. To verify the accuracy of the extracted area of interest, KOMPSAT-3A satellite images were used which taken on October 28, 2018 and July 7, 2018. The baseline samples for 2015 were extracted using the unchanged area of the continuous digital topographic map for 2013-2015 and the land cover map for 2015, and also extracted the baseline samples in 2018 using the unchanged area of the continuous digital topographic map for 2015-2017 and the land cover map for 2015. The redundant areas that occurred when merging continuous digital topographic maps and land cover maps were removed to prevent confusion of data. Finally, the checkpoints are generated within the region of interest, and the accuracy of the region of interest extracted from the K3A satellite images and the error matrix in 2015 and 2018 is shown, and the accuracy is approximately 93% and 72%, respectively. The accuracy of the region of interest can be used as a region of interest, and the misclassified region can be used as a reference for change detection.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.13 no.1
• /
• pp.147-155
• /
• 2010

Landcover information of access-denied area was extracted from low-medium and high resolution satellite image. Training for supervised classification was performed to refer visually by landcover map which is made and distributed from The Ministry of Environment. The classification result was compared by relating data of FACC land classification system. As we rasterize digital military map with same pixel size of satellite classification, the accuracy test was performed by image to image method. In vegetation case, ancillary data such as NDVI and image for seasons are going to improve accuracy. FACC code of FDB need to recognize the properties which can be automated.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.15 no.4
• /
• pp.26-41
• /
• 2012

It is necessary to manage, forecast and prepare agricultural production based on accurate and up-to-date information in order to cope with the climate change and its impacts such as global warming, floods and droughts. This study examined the applicability as well as challenges of the object-based image analysis method for developing a land cover image classification algorithm, which can support the fast thematic mapping of wide agricultural areas on a regional scale. In order to test the applicability of RapidEye's multi-temporal spectral information for differentiating agricultural land cover types, the integration of other GIS data was minimized. Under this circumstance, the land cover classification accuracy at the study area of Kimje ($1300km^2$) was 80.3%. The geometric resolution of RapidEye, 6.5m showed the possibility to derive the spatial features of agricultural land use generally cultivated on a small scale in Korea. The object-based image analysis method can realize the expert knowledge in various ways during the classification process, so that the application of spectral image information can be optimized. An additional advantage is that the already developed classification algorithm can be stored, edited with variables in detail with regard to analytical purpose, and may be applied to other images as well as other regions. However, the segmentation process, which is fundamental for the object-based image classification, often cannot be explained quantitatively. Therefore, it is necessary to draw the best results based on expert's empirical and scientific knowledge.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2012.05a
• /
• pp.750-750
• /
• 2012

토지이용의 변화는 수문순환의 변화를 초래한다. 이는 토지이용에 따른 지표 특성의 변화로 열수지가 변화함으로써 미규모 및 중규모 기상현상에 직접적으로 영향을 미치기 때문이다. 최근 4대강 살리기 사업 등의 대규모 토목사업으로 인해 넓은 면적의 지표면 상태가 변화하고 이에 따른 수문순환 변화에 대한 관심과 함께 연구가 진행되고 있다. 또한 최근 수문학 분야에서의 가장 큰 관심은 최신 기법을 적용한 수문분석 및 정확한 수문인자를 추출하고자 하는 것으로 원격탐사나 GIS 기법을 이용할 경우 유역의 물리적인 요소를 고려할 수 있으며 넓은 지역에 대한 정보를 공간적으로 분석할 수 있는 장점이 있다. 본 연구에서는 길이 103.5km, 유역면적 $1,478km^2$, 한강의 제 1지류인 섬강유역을 대상으로 위성 영상을 이용하여 토지피복에 따른 열수지 변화 특성을 모니터링 하였다. 대상지역의 토지피복은 Landsat TM 위성영상을 이용하여 수역, 도심지, 나지, 초지, 산림, 농경지, 구름 및 그림자 등으로 분류하였으며, 위성영상을 열수지 모델에 적용하여 인자별 격자기반의 공간분포도를 추출하였다. 또한 이를 이용하여 토지피복 종류에 따른 유역내 열환경 특성을 정량적으로 파악하고 열환경 관점에서 유리한 토지피복 조성을 도출하고자 한다.

• 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 the Korean Association of Geographic Information Studies
• /
• v.18 no.2
• /
• pp.135-148
• /
• 2015

This research is carried out for the land cover change detection in the Nakdong River basin before and after the 4 major rivers restoration project using the LiDAR DEM(Digital Elevation Model) and the multi-temporal Landsat imagery. Firstly the river basin polygon is generated by using the levee boundaries extracted from the LiDAR DEM, and the four river basin imagery are generated from the multi-temporal Landsat-5 TM(Thematic Mapper) and Landsat-8 OLI(Operational Land Imager) imagery by using the generated river basin polygon. Then the main land covers such as river, grass and bare soil are separately generated from the generated river basin imagery by using the image classification method, and the ratio of each land cover in the entire area is calculated. The calculated land cover changes show that the areas of grass and bare soil in the entire area have been significantly changed because of the seasonal change, while the area of the river has been significantly increased because of the increase of the water storage. This paper contributes to proposing an efficient methodology for the land cover change detection in the Nakdong River basin using the LiDAR DEM and the multi-temporal satellite imagery taken before and after the 4 major rivers restoration project.

• Korean Journal of Remote Sensing
• /
• v.27 no.4
• /
• pp.457-466
• /
• 2011

The objectives of this study are to produce level 3 type LULC map and analysis of phenological features of North Korea, ISODATA clustering of the 88scenes of MVC of MODIS NDVI in 2008 and 8scenes in 2009 was carried out. Analysis of phenological phases based mapping method was conducted, In level 2 type map, the confusion matrix was summarized and Kappa coefficient was calculated. Total of 27 typical habitat types that represent the dominant species or vegetation density that cover land surface of North Korea in 2008 were made. The total of 27 classes includes the 17 forest biotopes, 7 different croplands, 2 built up types and one water body. Dormancy phase of winter (${\sigma}^2$ = 0.348) and green up phase in spring (${\sigma}^2$ = 0.347) displays phenological dynamics when much vegetation growth changes take place. Overall accuracy is (851/955) 85.85% and Kappa coefficient is 0.84. Phenological phase based mapping method was possible to minimize classification error when analyzing the inaccessible land of North Korea.

• 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.

• The Journal of the Korea Contents Association
• /
• v.9 no.6
• /
• pp.436-443
• /
• 2009

There have been rapid increases to the demands for modeling diverse and complex spatial phenomena and utilizing spatial data through the computer across all the aspects of society. As a result, the importance and utilization of remote sensing and GIS's(geographic information systems) have also increased. It can produce digital data of enormous accuracy and value by incorporating remote sensing images into GIS analysis technology and make various thematic maps by classifying and analyzing land cover. Once such a map is made for the target area, it can easily do modeling and constant monitoring based on the map, revise the database with ease, and thus efficiently update geo-spatial information. Under the goal of analyzing changes to land cover along the road by combining the remote sensing and GIS technology, this study classified land cover from the images of two periods, detected changes to the six classes over ten years, and obtained statistics about the study area's quantitative area changes in order to provide basic decision making data for urban planning and development. By analyzing land use along the road, one can set up plans for the area along the road and the downtown to supplement each other.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.14 no.1
• /
• pp.39-47
• /
• 1996

This paper is to provide land cover classification techniques for large area ranged in different pathos by classifying Landsat TM data of Jeonnam province. The analyses proceeded by individual scene because acquired dates are not same in different pathes. In this processing, troubles had happened something like variation of classes can be classified in two scenes and choice problem about overlapped area. Since spatial effects in large area affect data values, it was difficult to make a selection of classes and training fields. we could present a solution about these problems by trial and error method, and found that Bayesian maximum likelihood classification and majority filtering were effective to improve classification accuracy.