A Method to Select Coherence Window Size for forest height estimation using PolInSAR Data

Abstract

Estimating the height of trees is one of the most important applications of polarimetrc interferometric synthetic aperture Radar (PolInSAR). PolInSAR requires an appropriate estimation of the interferogram coherence for obtaining the best results. Actually, the coherence estimation has a great impact on PolInSAR results for estimating the height of trees. Generally, the random volume over ground (RVOG) model is used for forest height estimation. In the ROVG, interferogram coherence entered as one of the observations. For coherence estimation, selecting the best window size conventionally is done by the coherence bias and convergence (CBC) method, which requires user experience and visual analysis. This study presents a fast and straightforward method to calculate the best value of the window size without affecting the user experience and visual analysis. In this study, we compare the performance of tree height estimation for the CBC and the proposed method on three simulated PolInSAR data in a forestry region. Experimental results obtained for estimating the height of trees show that, the window size values obtained by the proposed method are the same as those values obtained by the CBC method. Moreover, results suggest that the proposed method is faster and more stability than the CBC method.