Abstract: To overcome the limitations of finite aperture effect and operator non-orthogonality in the original dispersive Radon transform (DRT) method, a sparse dispersive Radon transform method is proposed to achieve more accurate guided wave mode separation and parameter estimation. Based on the sparsity of guided wave signal, the sparse DRT method introduces the orthogonal matching pursuit (OMP) algorithm to construct the cost function, and obtains the sparse energy representation in the DRT domain through multiple iterations, accurately separating the aliasing mode and denoising. For simulated signals with different signal-to- noise ratio (SNR), the signal reconstruction error of sparse DRT method is reduced by 51.1% compared with the original DRT method. Simulation and experiment demonstrate the feasibility of the proposed sparse DRT method for multimode separation and parameter estimation.