A general quantitative identification algorithm of subsurface defect for infrared thermography

Based on the Levenberg-Marquardt (L-M) method, a new general algorithm of estimating the size, depth and thermal conductivity of the subsurface defect according the surface temperature distribution obtained by passive, steady-state thermography is presented. The numerical experimental results show that the precise size, depth and the thermal conductivity of the defect can be estimated with or without the consideration of the temperature measurement error; and the larger the maximum temperature difference at the inspection surface, the higher the precision level of the identification results. The algorithm can be used to estimate the subsurface defect directly based on the thermal image or as part of the post-processing module in the infrared sensor system.