Abstract: In response to the challenges of low target contrast and significant background interference in the thermal imaging detection of tunnel water leakage,a thermal image grayscale gain mapping function and a crack identification and localization algorithm process are proposed.Firstly,the temperature field data of the leakage water lining is extracted,and a segmented mapping function is employed to compress the grayscale mapping of the background area,while stretching the grayscale mapping of the target area for improved recognizability.Subsequently,the optimal threshold range of the temperature field is determined,leading to the binarization of the infrared thermal image.Finally,the lining cracks are identified using the Canny edge detection algorithm,with the pixel coordinates of the target area being indexed using a traversal algorithm.Additionally,a thermal image detection experiment for simulated tunnel environment with leaking water and damaged lining is conducted,demonstrating that the proposed grayscale mapping function enhances the thermal image.The efficacy of this identification and localization method has been successfully applied to the lining damage detection of Jinan Kaiyuan Tunnel in China,providing a technical reference for similar tunnel engineering inspections.