A New Improved Algorithm Based on Three-Stage Inversion Procedure of Forest Height

Abstract

The three-stage inversion algorithm and the random volume over ground (RVoG) model can estimate forest parameters, using single baseline polarimetric synthetic aperture radar interferometry (PolInSAR) data. However, the classical method to invert forest height cannot hold the real data, as there are a lot of error data, which is not consistent with practice in theory. Accordingly, the results of classical inversion method are not accuracy and have many noise points. To overcome this problem, the improved algorithm eliminates the singular points while fitting the least square line to determine the location of the ground phase in complex unit circle. This new algorithm is more suitable for practical data.