Recently, there has been a significant interest in employing$nhigh frequency radar imagery for many industrial oriented and security$napplications. The quality control of aeronautics composite multi-layered$nmaterials and structures through Non-Destructive Testing (NDT) is the$nmain focus of this study. The main motivations for using this technology$nare: it allows penetration of most non-metal and non-polarized$nmaterials, it provides the ability for three-dimensional (3-D) imagery$nand in-depth information, and the millimeter waves (mmW) pose no health$nrisk to the operator. The main goal of this research is to develop an$nintegrated, semi-automatic and near real-time mode-operated image$nprocessing methodology for frequency-modulated continuous-wave (FMCW)$nmillimeter wave images with center frequencies around 100 GHz and 300$nGHz. The proposed methodology firstly foresees to process reflectance$nand transmittance 3-D imagery by extracting areas of interest and$nobject's boundaries at two operated frequencies. Then, the detected$nareas are subjected to a multi-source database and integrated by a$ndecision tree algorithm. The fused information is used to identify$ndefects and in-homogeneities within the objects. Finally, the$npost-processing phase examines and evaluates the spatial accuracy of the$nextracted information.