Soft defects localization by signature magnification with selective windowing

The SMSW (Signature Magnification by Selective Windowing) method, based on a temporal processing treat the reflectogram so as to make soft defect signatures detectable. The algorithm first performs a localization of critical points of the reflectogram (zero-crossing or mean-crossing). Then, it selects points having enough energy in both sides (above threshold). This threshold is to be determined statistically or set depending on the application. Once the defect area is windowed, we proceed to the magnification procedure which amplify the defect signature while reducing the noise level on the reflectogram. The proposed method shows very convincing results both in simulation and experiments and for different types of cable and transmission line. We have demonstrated and assessed its applicability on coaxial cables, twisted pair and micro-strip lines. We found that we achieve a gain even with low SNR (Signal to Noise Ratio).