基于红外图像增强的衬砌裂损渗漏水识别方法及工程应用

Method for Identifying Leakage Water in the Tunnel Lining Based on Infrared Image Enhancement and Engineering Application

针对隧道渗漏水红外热成像检测存在目标对比度小、背景干扰严重的问题,提出了红外图像灰度增益映射函数及裂损识别定位算法流程.首先,提取渗漏水衬砌的温度场数据,通过分段式映射函数压缩背景区的灰度映射,拉伸识别目标区的灰度映射,提高识别目标区的可辨识度.然后,求解温度场最佳阈值范围,将红外热图像二值化.最后,通过Canny边缘检测算法识别衬砌裂损形式.利用遍历算法索引目标区像素坐标位置.此外,开展了模拟隧道环境的裂损衬砌渗漏水红外检测实验分析,结果表明,提出的灰度映射函数可以增益红外热图像.该识别定位方法已经成功应用于济南开元隧道的衬砌病害检测,可为类似隧道工程检测提供技术参考.

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.