TY - JOUR
T1 - Above-Bottom Biomass Retrieval of Aquatic Plants with Regression Models and SfM Data acquired by a UAV Platform – A Case Study in Wild Duck Lake Wetland, Beijing, China
AU - Jing, Ran
AU - Gong, Zhaoning
AU - Zhao, Wenji
AU - Pu, Ruiliang
AU - Deng, Lei
PY - 2017/12/1
Y1 - 2017/12/1
N2 - Above-bottom biomass (ABB) is considered as an important parameter for measuring the growth status of aquatic plants, and is of great significance for assessing health status of wetland ecosystems . In this study, Structure from Motion (SfM) technique was used to rebuild the study area with high overlapped images acquired by an unmanned aerial vehicle (UAV). We generated orthoimages and SfM dense point cloud data, from which vegetation indices (VIs) and SfM point cloud variables including average height (HAVG), standard deviation of height (HSD) and coefficient of variation of height (HCV) were extracted. These VIs and SfM point cloud variables could effectively characterize the growth status of aquatic plants, and thus they could be used to develop a simple linear regression model (SLR) and a stepwise linear regression model (SWL) with field measured ABB samples of aquatic plants. We also utilized a decision tree method to discriminate different types of aquatic plants. The experimental results indicated that (1) the SfM technique could effectively process high overlapped UAV images and thus be suitable for the reconstruction of fine texture feature of aquatic plant canopy structure; and (2) an SWL model based on point cloud variables: HAVG, HSD, HCV and two VIs: NGRDI, ExGR as independent variables has produced the best predictive result of ABB of aquatic plants in the study area, with a coefficient of determination of 0.84 and a relative root mean square error of 7.13%. In this analysis, a novel method for the quantitative inversion of a growth parameter (i.e., ABB) of aquatic plants in wetlands was demonstrated.
AB - Above-bottom biomass (ABB) is considered as an important parameter for measuring the growth status of aquatic plants, and is of great significance for assessing health status of wetland ecosystems . In this study, Structure from Motion (SfM) technique was used to rebuild the study area with high overlapped images acquired by an unmanned aerial vehicle (UAV). We generated orthoimages and SfM dense point cloud data, from which vegetation indices (VIs) and SfM point cloud variables including average height (HAVG), standard deviation of height (HSD) and coefficient of variation of height (HCV) were extracted. These VIs and SfM point cloud variables could effectively characterize the growth status of aquatic plants, and thus they could be used to develop a simple linear regression model (SLR) and a stepwise linear regression model (SWL) with field measured ABB samples of aquatic plants. We also utilized a decision tree method to discriminate different types of aquatic plants. The experimental results indicated that (1) the SfM technique could effectively process high overlapped UAV images and thus be suitable for the reconstruction of fine texture feature of aquatic plant canopy structure; and (2) an SWL model based on point cloud variables: HAVG, HSD, HCV and two VIs: NGRDI, ExGR as independent variables has produced the best predictive result of ABB of aquatic plants in the study area, with a coefficient of determination of 0.84 and a relative root mean square error of 7.13%. In this analysis, a novel method for the quantitative inversion of a growth parameter (i.e., ABB) of aquatic plants in wetlands was demonstrated.
KW - Wetland aquatic plants
KW - Above-bottom biomass
KW - Unmanned aerial vehicle
KW - SfM data
KW - Regression analysis
UR - https://digitalcommons.usf.edu/geo_facpub/1357
UR - https://doi.org/10.1016/j.isprsjprs.2017.11.002
U2 - 10.1016/j.isprsjprs.2017.11.002
DO - 10.1016/j.isprsjprs.2017.11.002
M3 - Article
VL - 134
JO - ISPRS Journal of Photogrammetry and Remote Sensing
JF - ISPRS Journal of Photogrammetry and Remote Sensing
ER -