For vegetation parameter inversion, the single baseline polarimetric SAR interferometry technique, such as the three-stage method and the ESPRIT algorithm, is limited by the observed data with the minimum ground to volume amplitude ratio, which effects the estimation of the effective phase center for the vegetation canopy or the surface, and thus results in the underestimated vegetation height. In order to remove this defect of the single baseline inversion techniques in some extend, another baseline polarimetric SAR interferometry data is added on vegetation parameter estimation in this paper, and a dual baseline polarimetric SAR interferometry technique for the extraction of the vegetation parameter is investigated and improved to reduce the dynamic bias for the vegetation parameter estimation. Finally, the simulated data and real data are used to validate this dual baseline technique.