• Korean Journal of Remote Sensing
• /
• v.39 no.5_4
• /
• pp.1145-1153
• /
• 2023

Urban trees play an important role in absorbing carbon dioxide from the atmosphere, improving air quality, mitigating the urban heat island effect, and providing ecosystem services. To effectively manage and conserve urban trees, accurate spatial information on their location, condition, species, and population is needed. In this study, we propose an algorithm that uses a high-resolution urban tree cover map constructed from deep learning approach to separate trees from the urban land surface and accurately detect tree locations through local maximum filtering. Instead of using a uniform filter size, we improved the tree detection performance by selecting the appropriate filter size according to the tree height in consideration of various urban growth environments. The research output, the location and height of individual trees in human settlement over Suwon, will serve as a basis for sustainable management of urban ecosystems and carbon reduction measures.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_2
• /
• pp.1633-1641
• /
• 2022

Urban population concentration and indiscriminate development are causing various environmental problems such as air pollution and heat island phenomena, and causing human resources to deteriorate the damage caused by natural disasters. Urban trees have been proposed as a solution to these urban problems, and actually play an important role, such as providing environmental improvement functions. Accordingly, quantitative measurement and analysis of individual trees in urban trees are required to understand the effect of trees on the urban environment. However, the complexity and diversity of urban trees have a problem of lowering the accuracy of single tree detection. Therefore, we conducted a study to effectively detect trees in Dongjak-gu using high-resolution aerial images that enable effective detection of tree objects and You Only Look Once Version 5 (YOLOv5), which showed excellent performance in object detection. Labeling guidelines for the construction of tree AI learning datasets were generated, and box annotation was performed on Dongjak-gu trees based on this. We tested various scale YOLOv5 models from the constructed dataset and adopted the optimal model to perform more efficient urban tree detection, resulting in significant results of mean Average Precision (mAP) 0.663.

• Journal of Environmental Impact Assessment
• /
• v.30 no.3
• /
• pp.155-163
• /
• 2021

The number of trees to be removed trees (ART) in the environmental impact assessment is an environmental indicator used in various parts such as greenhouse gas emissions and waste of forest trees calculation. Until now, the ART has depended on the forest tree density of the vegetation survey, and the uncertainty of estimating the amount of removed trees has increased due to the sampling bias. A full-scale survey can be offered as an alternative to improve the accuracy of ART, but the reality is that it is impossible. As an alternative, there is an individual tree detection using aerial image (ITD), and in this study, we compared the ARTs estimated by full-scale survey, sample survey, and ITD. According to the research results, compared to the result of full-scale survey, the result of ITD was overestimated by 25. While 58 were overestimated by the sample survey (average). However, as the sample survey is an estimate based on random samples, ART will be overestimated or underestimated depending on the number and size of quadrats.

• Spatial Information Research
• /
• v.20 no.3
• /
• pp.27-38
• /
• 2012

As the necessity of forest conservation and management has been increased, various forest studies using LIDAR data have been actively performed. These studies often utilize the tree height as an important parameter to measure the forest quantitatively. This study thus attempt to apply two representative methods to estimate tree height from airborne LIDAR data and compare the results. The first method based on the detection of the individual trees using a local maximum filter estimates the number of trees, the position and heights of the individual trees, and the mean tree height. The other method estimates the maximum and mean tree height, and the crown mean height for each grid cell or the entire area from the canopy height model (CHM) and height histogram. In comparison with the field measurements, 76.6% of the individual trees are detected correctly; and the estimated heights of all trees and only conifer trees show the RMSE of 1.91m and 0.75m, respectively. The tree mean heights estimated from CHM retain about 1~2m RMSE, and the histogram method underestimates the tree mean height with about 0.6m. For more accurate derivation of diverse forest information, we should select and integrate the complimentary methods appropriate to the tree types and estimation parameters.

• Korean Journal of Remote Sensing
• /
• v.39 no.6_3
• /
• pp.1721-1730
• /
• 2023

As awareness of the problem of global warming emerges around the world, the role of carbon sinks in settlement is increasingly emphasized to achieve carbon neutrality in urban areas. In order to manage carbon sinks in settlement, it is necessary to identify the current status of carbon sinks. Identifying the status of carbon sinks requires a lot of manpower and time and a corresponding budget. Therefore, in this study, a map predicting the location of trees was created using already established tree location information and Sentinel-2 satellite images targeting Seoul. To this end, after constructing a tree presence/absence dataset, structured data was generated using 16 types of vegetation indices information constructed from satellite images. After learning this by applying the Extreme Gradient Boosting (XGBoost) model, a tree prediction map was created. Afterward, the correlation between independent and dependent variables was investigated in model learning using the Shapely value of Shapley Additive exPlanations(SHAP). A comparative analysis was performed between maps produced for local parts of Seoul and sub-categorized land cover maps. In the case of the tree prediction model produced in this study, it was confirmed that even hard-to-detect street trees around the main street were predicted as trees.

• Proceedings of the Korean Association of Geographic Inforamtion Studies Conference
• /
• 2008.06a
• /
• pp.260-264
• /
• 2008

변화탐지는 도시모델의 갱신을 위해 중요한 단계이다. 이에 본 연구는 도시지역의 변화탐지를 위한 라이다데이터로부터 추출한 표면패치의 분류 방법을 제안한다. 제안된 방법의 주요 과정은 (1) 라이다 데이터로부터 생성된 DSM의 차분을 통해 변화영역을 탐지하고, (2) 탐지된 영역의 라이다 점으로부터 표면패치를 구성하고, (3) 구성된 각각의 패치의 종류를 지면 수목, 빌딩으로 분류한다. 제안된 방법을 실측데이터에 적용한 결과를 동일한 지역의 정사영상으로부터 육안검사를 통해 수동 생성된 기준데이터를 이용하여 검증하였다. 패치분류의 성공률은 99%로 평가되었다. 결론적으로 제안된 방법은 변화탐지를 위한 강인하고, 신뢰성이 높고, 효율적인 패치 분류방법으로 판단된다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2012.11a
• /
• pp.516-518
• /
• 2012

일반적으로 표적에 대한 탐지는 감시장비의 성능과 지형지물의 차폐 여부가 가장 큰 영향을 준다. 본 연구는 SRTM DSM (Digital Surface Model)과 국방지형정보단 DEM (Digital Elevation Model) 그리고 여기에 수목고를 고려한 DCM (Digital Canopy Model)고도를 기반으로 탐지확률 실험을 하였다. 실험결과 DCM과 DEM 기반의 탐지확률 결과가 가장 유사성이 높았고, SRTM과 DEM 기반의 탐지 확률은 차이가 나는 것으로 확인하였다. 따라서 SRTM이 이론적으로 DSM으로 고려되지만, 향후 추가적인 연구를 통해 이에 대한 분석이 더 필요할 것으로 사료된다.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.30 no.3
• /
• pp.259-268
• /
• 2012

This paper aims to estimate the carbon dioxide stocks in forests using airborne LiDAR data with a density of approximate 4.4 points per meter square. To achieve this goal, a processing chain consisting of bare earth Digital Terrain Model(DTM) extraction and individual tree top detection has been developed. As results of this experiment, the reliable DTM with type-II errors of 3.32% and tree positions with overall accuracy of 66.26% were extracted in the study area. The total estimated carbon dioxide stocks in the study area using extracted 3-D forests structures well suited with the traditional method by field measurements upto 7.2% error level. This results showed that LiDAR technology is highly valuable for replacing the existing forest resources inventory.

• Korean Journal of Environmental Biology
• /
• v.41 no.1
• /
• pp.18-30
• /
• 2023

Forest destruction is an inevitable result of the development processes. According to the environmental impact assessment, over 10% of the destroyed trees need to be recycled and transplanted to minimize the impact of forest destruction. However, the rate of successful transplantation is low, leading to a high rate of tree death. This is attributable to a lack of consideration for environmental factors when choosing a temporary site for transplantation and inadequate management. To monitor transplanted trees, a field survey is essential; however, the spatio-temporal aspect is limited. This study evaluated the applicability of remote sensing for the effective monitoring of transplanted trees. Vegetation indices based on satellite remote sensing were derived to detect time-series changes in the status of the transplanted trees at three temporary transplantation sites. The mortality rate and vitality of transplanted trees before and after the transplant have a similar tendency to the changes in the vegetation indicators. The findings of this study showed that vegetation indices increased after transplantation of trees and decreased as the death rate increased and vitality decreased over time. This study presents a method for assessing newly transplanted trees using satellite images. The approach of utilizing satellite photos and the vegetation index is expected to detect changes in trees that have been transplanted across the country and help to manage tree transplantation for the environmental impact assessment.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.26 no.4
• /
• pp.341-350
• /
• 2008

Change detection has been recognized as one of the most important steps to update city models. In this study, we thus propose a method to detect urban changes from two sets of LIDAR data acquired at different times. The main processes in the proposed method are (1) detecting change areas through subtraction between two DSMs generated from the LIDAR sets, (2) organizing the LIDAR points within the detected areas into surface patches, (3) classifying the class of each patch such as ground, vegetation, and building, and (4) determining the kinds of changes based on the properties and classes of the patches. The results which were obtained from the application of the proposed method to real data were verified as appropriate using the reference data manually acquired from the visual inspection of the orthoimages of the same area. The probability of success in change detection is assessed to 97% on an average. In conclusion, the proposed method is evaluated as a reliable, and efficient approach to change detection and thus the update of city model.