Traditional gamma safeguards measurements have usually been performed using a segmented gamma scanning (SGS) system. The accuracy of this technique relies on the assumption that the sample matrix and the activity are both uniform for a segment. Waste barrels are often highly heterogeneous, span a wide range of composition and matrix type. The primary sources of error are all directly or indirectly related to a non-uniform measurement response associated with unknown radioactive source spatial distribution and heterogeneity of the matrix. These errors can be significantly reduced by some imaging techniques that measure exact spatial locations of sources and attenuation maps. In this paper we describe a joint R&D effort between the Lawrence Livermore National Laboratory (LLNL) and the Institute of Nuclear Techniques (INT) of the Technical University, Budapest, to compare results obtained by two different gamma-ray nondestructive assay (NDA) systems used for imaging waste barrels. The basic principles are the same, but the approaches are different. Key factors to judge the adequacy of a method are the detection limit and the accuracy. Test drums representing waste to be measured are used to determine basic parameters of these techniques.