2026 年 75 巻 9 号 p. 394-399
This study introduces a remote, non-contact crack inspection technique for steel structures using an active temperature gap method combined with infrared thermography. A concentrated mirror halogen heating system was developed to enable localized heating from a distance without physical access. Temperature responses during heating and cooling were analyzed and validated against theoretical predictions. Crack detection was achieved by visualizing thermal anomalies caused by heat transfer interruption at crack locations. To enhance detection accuracy, discrete Fourier transform (DFT) analysis was applied to time-series thermal data, improving the signal-to-noise ratio and highlighting crack features. Additionally, Sobel filtering was employed for edge extraction and accurate crack length estimation. The proposed method eliminates the need for scaffolding, significantly improves inspection efficiency, and offers strong potential for automation and large-scale applications in bridges and industrial plants. This approach provides a practical solution for structural health monitoring, reducing labor costs and safety risks associated with conventional inspection methods.