Estimation of Broadleaf Tree Canopy Height of Wolong Nature Reserve Based on InSAR and Machine Learning Methods

Tree height is an important parameter for calculating forest carbon sink and assessing cycle. In order to obtain tree over a large area both efficiently at low cost, this study proposed Interferometric Synthetic Aperture Radar (InSAR) combined with machine learning method estimate the canopy height. The in was obtained using Unmanned Aerial Vehicle (UAV) photogrammetry, which considered be true Two methods (Random Forest, Multi-layer perceptron) were used establish relationship between phase center calculated by InSAR DEM differential interference coherent amplitude topographic factor, backward scattering coefficient coherence introduced into model. It found that accuracy of estimation random two can reach 0.95 0.94. root-mean-square error 1.76 m, 1.86 respectively. multi-layer perceptron 0.25 0.2. 3.96 m 4.13 m. results indicated combination cost. Moreover, integrated algorithm demonstrated better stability higher than single perceptron.