Enhanced Infrared Detection of Micro-Leakage in Utility Tunnel Linings via Local Temperature Gradient Mapping

Underground utility tunnels in service are prone to micro-cracks that cause subtle water leakage, which is difficult to detect manually. Infrared thermography offers non-contact potential but suffers from low contrast and background interference when identifying tiny leakage spots. This study focuses on punctate micro-leakage spots in concrete linings as the direct leakage source and establishes a laboratory simulation system to investigate temperature evolution during seepage. By analyzing the spatiotemporal characteristics of the temperature field, a novel infrared image enhancement method is proposed: it extracts the local temperature core region around the defect, computes a gradient gain factor based on the temperature variation within that region, and applies it to grayscale mapping of the global thermal image. This approach significantly improves the contrast between leakage areas and the background, enabling accurate identification and precise localization of point-like leakage defects. Experimental results on a concrete specimen with a simulated crack validate the method’s effectiveness under various water temperatures and flow rates. The proposed technique provides practical support for infrared-based leakage hazard detection in underground utility tunnels.