A methodology for identification of multiple objects using their natural resonant frequencies from both time and frequency domain data is presented. This methodology is applied to generate a library of poles of various objects in the frequency domain and compare them with the computed natural poles of unknown objects using the typical transient temporal late time response in the time domain. The Cauchy method is applied directly in the frequency domain to extract the Singularity Expansion Method (SEM) poles and thus a library of poles of various perfectly conducting objects (PEC) is generated. The Matrix Pencil (MP) method is then applied to the late time response to compute the SEM poles for identification in the time domain. Simulation examples are analyzed to illustrate the potential of this proposed methodology.