To provide short-range super-resolution ultra-wideband (UWB) radar under multi-scattering conditions, a superior wavefront extraction algorithm is proposed in this paper. Conventional correlation-based pulse separation methods based on SAGE, CLEAN or the previously introduced superior dynamic correlation method (DCM) are revised, validated, and compared. In this paper, the DCM is improved significantly by applying the Pauli scattering matrix decomposition onto the radar data. This novel wavefront extraction algorithm is called polarimetric dynamic correlation method (PDCM) and is suitable to resolve several overlapping pulses, which consist of both strong echoes and weak echoes which are masked by the strong ones. The performance of the PDCM and the comparison with alternative algorithms is carried out by a subsequent feature extraction algorithm for visual verification. Experimental validations are performed with two complex test objects, a maximum length sequence radar device (4.5–13.5 GHz) and compact dual-polarized UWB antennas.