A spatio-temporal fusion method for non-destructive testing using infrared thermography

Abstract

The infrared thermography, as a non-contact and highly sensitive non-destructive testing (NDT) method, has been extensively applied in various fields such as aerospace, construction, and industrial manufacturing industry. Current post-processing of dynamic thermal sequences relies mainly on thermal images’ spatial information without incorporating temporal data, compromising defect detection accuracy and efficiency. To address this challenge, we propose a spatio-temporal fusion thermography (STFT) method that enables high-precision of surface defects detection. The method primarily establishes a theoretical model by integrating spatial and temporal gradient information from dynamic thermal sequences, with thermal propagation being estimated using optical flow method. The verification experiment on metal surface defects indicates that the STFT method is capable of reliably detecting surface microcracks as small as 3 μm in width. And it can effectively eliminate the impact of uneven temperature distribution on the test specimens and significantly improves the signal-to-noise ratio (SNR) of the defect images. The proposed method holds great potential for broad application in the field of industrial non-destructive testing, such as detecting surface or near surfac cracks and pitting defects in gears.