Research on infrared feature recognition algorithm for water leakage in highway tunnels

With the rapid development of the world economy, the number and scale of road transportation infrastructure have grown rapidly. In recent years, a large number of highway tunnels have been put into use. The prevalence of water seepage diseases in road tunnels poses a major hazard to the safety of personnel and vehicles. Water seepage detection has been one of the high-priority research topics for tunnel inspection. Previous studies have mainly used a single visible camera as a sensor to detect water leakage areas in images, which is extremely sensitive to insufficient lighting in the tunnel. Considering the existence of temperature difference between the water leakage area and the normal area, we used an infrared camera to collect tunnel lining images and constructed a water leakage area identification algorithm based on the temperature difference between the water leakage area and the normal area using K-means clustering algorithm and low-pass filtering algorithm based on the temperature field distribution of infrared images. This study can provide a new idea for the rapid identification of tunnel water leakage.