Data Fusion for Enhanced Defect Detectability in Non-Stationary Thermal Wave Imaging

Subsurface analysis to determine the location and the sizing of an anomaly using thermography opened avenues for the investigation of various stimulation mechanisms and processing approaches. Non-stationary thermal wave imaging caters to it by employing a suitable band of frequencies according to the depth resolution of interest at low peak powers as supported by various processing approaches. But no single method is capable to explore all the details underneath the test object surface and provide subsequent details. In order to overcome it and to embed a variety of details obtained from various processing approaches, this letter introduces fusion-based methodology by combining the details obtained from principal component and other contemporary processing approaches for quadratic frequency modulated thermal wave imaging through the experimentation carried over a carbon fiber reinforced polymer specimen.