선택한 단어 수는 입니다.
최소 단어 이상 선택하여야 합니다.
최소 단어 이상 선택하여야 합니다.
선택한 단어 수는 30입니다.
최대 10 단어까지만 선택 가능합니다.
최대 10 단어까지만 선택 가능합니다.
문제 정의
- , 2012) due to its unique ability to capture vertical structural information, little research has focused on the utilization of LiDAR data for mapping urban green infrastructure in urban environments. As a preliminary step for mapping green infrastructure utilizing multi-source remote sensing data in urban environments, the objective of this study is to map vegetation volume by fusing LiDAR and multispectral data in urban environments. Multispectral imageries are used to identify the two dimensional distribution of green infrastructure, while LiDAR data are utilized to characterize the vertical structure of the identified green structure.
제안 방법
- 2 illustrates flow charts for calculating vegetation volume by fusing LiDAR and multispectral data. First, the DHM layer was generated from the LiDAR data, and the NDVI layer was calculated from the ortho imageries. Second, the DHM and the NDVI layers were fused at the pixel level.
- 85 points/pixel). LiDAR point cloud data were classified into ground and non-ground classes using LAStools (Isenburg, http://lastools.org), and the ground points were used to generate 5 foot digital terrain model (DTM) using the natural neighbor interpolation algorithm. Digital surface model (DSM) was calculated by calculating maximum elevation within each grid cell.
- LiDAR is an active remote sensing technique that enables three dimensional measurements of objects on the Earth’s surface based on range measurements. The range measurements are accomplished by measuring time difference between outgoing laser pulse and return laser pulse. The calculated range measurements are combined with location of the sensor computed from the GPS (Global Positioning System) and attitude of the sensor estimated from the IMU (Inertial Measurement Unit) sensor, so that three dimensional coordinates of the objects are calculated through a vector addition operation.
- Due to the different projection used in the vegetation volume layer and the NLCD 2006 data, coordinates of the vegetation volume layer is converted into the Albers Conical Equal Area projection to co-register both dataset. Then, the 1 arc-second grid structure of the NLCD 2006 data was placed on top of the converted vegetation volume layer and vegetation volume of each land cover was calculated by summing up the vegetation volume layer pixels within each NLCD 2006 grid cell.
- The experimental results indicated that vegetation volume varies greatly even in the same land use cover, and the traditional land cover map based above ground biomass estimation approach may introduce bias in the estimation results since the traditional approach lacks the ability to capture this variation that exists especially in the developed land cover classes. Unfortunately, this study adopted the simplest vegetation volume estimation model with an assumption that the amount of vegetation is linearly correlated with the vegetation height, and this model does not take into account vertical structural differences that exist even in the same land cover class. Future work will include developing better estimation models that are based on field measurement data and taking advantages of vertical structural information extracted from the LiDAR data.
- Multispectral imageries were used as mask to filter out urban objects such as buildings and roads, and LiDAR data were used to calculate volume of vegetated area. Vegetation volumes of the 16 classes were calculated by overlaying the NLCD2006 grid structure over the vegetation volume layer and summing up vegetation volume within the each grid cell. The experimental results indicated that Forested classes (DF, EF, MF, and WW) forms a group which has the highest average vegetation volume, and they also showed highest vegetation volume variation within them.
대상 데이터
- A topographic LIDAR mapping system with the Leica Geosystem ALS50 (Airborne Laser Scanner 50) sensors were flown over the study area to collect LIDAR data during two different time periods: November 21-29, 2012 and April 11-14 2009, and this campaign consists of a total of 167 flight lines. The ALS50 is a discrete return LiDAR system, and can operate up to 150 KHz pulse repetition frequency (PRF) rate and 75 degree field of view (FOV).
- This study is conducted in the Cook County that is located in the north eastern part of Illinois, USA (Fig. 1). The Cook County encompasses the metropolitan Chicago city area which has a population of 2,695,598 and the Chicago city is the third largest city in the United States (U.
성능/효과
- This group of land covers contains smaller vegetation volume than the forested land covers since the SS land cover is expected to contain mostly shrubs that are much shorter than trees and much smaller number of canopies than the forested land cover group. The experimental results that the DOS and DLI land covers belongs to this group and they contain significant amount of vegetation volume may be explained by the fact that 1) residential area is one of the major components of the DLI land cover and trees are usually planted around the residential area, and 2) the DOS land cover is usually co-located around the developed areas and used as ecosystem service area with vegetation where residents can relax and entertain themselves. Third, a group of land covers including Open Water (OW), Developed Medium Intensity (DMI), Grassland Herbaceous (GH), Pasture Hay (PH), and Emergent Herbaceous Wetlands (EHW) forms a cluster which has third highest vegetation volume ( ~ 1,000 m3/pixel).
- In the traditional approach to map urban forest using optical remote sensing data, developed land cover classes are usually not included in the analysis since little research has been done in urban environments and no factor is available for these land cover classes. However, the experimental results indicate that developed land cover classes (DOS, DLI, DMI, DHI) contain major portion (approximately 76.6%) of total vegetation volume in urban environments and the developed land cover classes should not be ignored in the carbon storage analysis.
- Although area percentages of the WW and DF land covers are less than 8%, they contain about 28% of total vegetation volume. The experimental results also indicate that quite amount of vegetation volume (11.1%) is contained in the DHI land cover class due to its larger coverage (13.7% of total area) even though the DHI land cover belongs to a group with the smallest vegetation volume. In the traditional approach to map urban forest using optical remote sensing data, developed land cover classes are usually not included in the analysis since little research has been done in urban environments and no factor is available for these land cover classes.
- The experimental results indicated that Forested classes (DF, EF, MF, and WW) forms a group which has the highest average vegetation volume, and they also showed highest vegetation volume variation within them. The experimental results also showed that DOS and DLI classes belongs to a group which has second highest average vegetation volume, but they contains the most vegetation volume among all classes due to their dominant coverage within the study area. The experimental results indicated that vegetation volume varies greatly even in the same land use cover, and the traditional land cover map based above ground biomass estimation approach may introduce bias in the estimation results since the traditional approach lacks the ability to capture this variation that exists especially in the developed land cover classes.
- Vegetation volumes of the 16 classes were calculated by overlaying the NLCD2006 grid structure over the vegetation volume layer and summing up vegetation volume within the each grid cell. The experimental results indicated that Forested classes (DF, EF, MF, and WW) forms a group which has the highest average vegetation volume, and they also showed highest vegetation volume variation within them. The experimental results also showed that DOS and DLI classes belongs to a group which has second highest average vegetation volume, but they contains the most vegetation volume among all classes due to their dominant coverage within the study area.
- The experimental results that the DOS and DLI land covers belongs to this group and they contain significant amount of vegetation volume may be explained by the fact that 1) residential area is one of the major components of the DLI land cover and trees are usually planted around the residential area, and 2) the DOS land cover is usually co-located around the developed areas and used as ecosystem service area with vegetation where residents can relax and entertain themselves. Third, a group of land covers including Open Water (OW), Developed Medium Intensity (DMI), Grassland Herbaceous (GH), Pasture Hay (PH), and Emergent Herbaceous Wetlands (EHW) forms a cluster which has third highest vegetation volume ( ~ 1,000 m3/pixel). Last, a group of land covers including Developed High Intensity (DHI), Barren Land (BL), and Cultivated Crops (CC) has the smallest vegetation volume (< 1,000 m3/pixel).
후속연구
- Unfortunately, this study adopted the simplest vegetation volume estimation model with an assumption that the amount of vegetation is linearly correlated with the vegetation height, and this model does not take into account vertical structural differences that exist even in the same land cover class. Future work will include developing better estimation models that are based on field measurement data and taking advantages of vertical structural information extracted from the LiDAR data.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.