Fusion of polarization degree and thermal parameters for enhanced nondestructive testing of coated metals

Accurate detection of the state of metal beneath a coating is essential for eliminating safety hazards arising from metal damage and corrosion. However, conventional visual inspection methods face challenges in extracting the characteristics of the base metal due to changes in the coating's color, roughness, polarization, and surface attributes. This paper presents a non-destructive testing method for assessing the metal state under coatings by fusing polarization and thermal parameters. We built a coated metal infrared polarization image acquisition platform and used it to capture polarization images and thermal diffusion infrared sequence diagrams in multiple sets of experiments on coated metals. Based on the polarization and infrared datasets of these coated metals, we develop a polarization data model and a thermal diffusion physical model for coated metals to extract relevant polarization and thermal related parameters, followed by a fusion-based classification of these modal parameters. Experimental results demonstrate the robustness of the proposed method in identifying the metal state under coatings, yielding more accurate classification results compared to using individual modal parameters alone. Our findings introduce a novel approach for non-destructive testing of metal beneath coatings, offering potential advancements in industrial inspection practices.