Abstract: For a model derived from the Geometrical Theory of Diffraction (GTD), a fast method based on spatial filtering is proposed to extract parameters of two-dimensional scattering centers. The proposed method utilizes spatial filtering process to decompose two-dimensional scattering centers extraction into several times of one-dimensional scattering centers extraction, in which the one-dimensional Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT) is employed to estimate the parameters of scattering centers for each dimensional independently. Finally, the pair-matching of two-dimensional parameters is accomplished by searching the minimums of Euclidean distance. Compared with the method based on two-dimensional ESPRIT, the proposed method does not need high-dimensional eigenvalue decomposition, thus the computational complexity is significantly reduced. Simulation results show that the proposed method not only reduces greatly the computational burden, but also keeps high accuracy of parameter estimation compared with 2D-ESPRIT algorithm, and it is proved to be effective in scattering center extraction.