도시 녹지공간 식생 모니터링을 위한 무인항공기 활용방안

Application of UAV for Vegetation Monitoring in Urban Green Space

Abstract

With the diversification of research using UAV(Unmanned Aerial Vehicle)s, the possibility of remote sensing research for urban green spaces is increasing. UAVs can be used as an investigation method to monitor the successful construction of the park and the planting of vegetation since its creation. This study was carried out to investigate UAVs utilization of urban green space monitoring in Dosol Square. It was photographed three times on May 21, July 13, and September 16, 2018 using DJI Phantom3 pro, Inspire2, and Parrot Sequoia multispectral camera. Orthographic images were overlaid on the planting plan of the site and the construction results were checked, the change of vitality of the plantation area was analyzed by NDVI(Normalized Difference Vegetation Index) and SAVI(Soil Adjusted Vegetation Index). As a result, it was confirmed that the UAVs are very effective for surveying the view of the urban green space after the construction and recording the results, which can be grasped quantitatively by overlaying the planting plan map. UAVs are more likely to be used in terms of monitoring vegetation vitality. It is interpreted that SAVI is better than NDVI in the green space just after composition. Chionanthus retusus and Pinus strobus were analyzed for their low level of vitality, and partially damaged and their vitality was lowered. In addition, there was difficulty in grass planting area and flower garden due to drainage and summer drought problems. In the future, it is expected that orthoimage and multispectral data using UAVs will be useful in the early vegetation monitoring and management field of urban green spaces.

Figure 1. Study area

Figure 2. Four orthoimages of a site classified by multispectral imaging(A : Green, B : Red, C : Red edge, D : NIR spectral bands)

Figure 3. UAV used for multispectral imageacquisition(A: RGB Camera; B: multispectral Camera;C: Solar light sensor)

Figure 5. Overlay analysis of aerial Photograph and plan drawing of study area

Figure 4. Calibration target for radial correction

Figure 6. Land cover changes during the two periods(2009 and 2017). Daum map image

Figure 7. Method of individual tree extraction in study area(A. light circles), and changes in Vegetation Indexes (B. NDVI, C. SAVI) by Three Periods(23 May, 13 July, 16 Sep.)

Figure 8. Comparison of Vegetation Index(NDVI, SAVI) mean value in three periods. (A : Zelkova serrata, B: Quercus palustris C : Chionanthus retusus, D : Grass Land, E : Pinus densiflora, F : Pinus strobus, G : Metasequoia glyptostroboides, H: Flower garden)

Table 1. Features of Photography

Table 2. Result of analysis that mean and maximum values ​of vegetation index according to species