An ISAR image is a synthetic picture of a radar target generated using scattered field data rather than direct observation of the object. It permits human visual recognition and identification of the object and shows a pictorial outline of the object which carries detailed information about the scattering centers of the target. ISAR image reconstruction basically depends on the extraction of scattering centers of the scattering object. The extraction of scattering centers can be accomplished by different techniques, which fall into one of two large categories: Fourier transform and parametric modeling based techniques. The purpose of this study is to examine one Fourier transform based and three parametric model based techniques, namely namely Prony's method, total least squares (TLS) Prony method and the generalized pencil of functions (GPOF) method and compare them on the basis of accuracy and computational efficiency. The use of the GPOF method in the ISAR image reconstruction is the original contribution of this study.