Clutter mitigation based of adaptive singular value decomposition in tomographic radar images for material inspection

To determine the quality and mechanical stability of composite materials, a defect detection whilst in the processing chain as well as active monitoring during use is inevitable. This can be achieved via the application of millimeter-wave radars, with the tomographic information extractable from the reflected signal. Due to clutter effects present in the reflected signal, the extracted tomographic image quality diminishes. Therefore, a method to reduce the impact of these clutter sources in the post-processing of the measurement while retaining the information of possible structure defects is presented. The method is based on singular value decomposition with the singular values determined through the comparison of the measured reflected signal with an ideal signal reference. With this method, the singular values do not have to be defined by hand, which represents a major advantage when considering automatic data evaluation possibilities.