English：

Crack detection is a critical component of operational maintenance for existing tunnels. Toenhance crack detection efficiency, a tunnel lining crack detection system based on image processingtechnology and computer vision has been proposed. This system comprises a video image acquisitiondevice, image stitching software, and crack detection algorithms. The video image acquisition devicecomprises six cameras and three lighting units mounted on the steel frame of an inspection vehicle. It canbe adjusted circumferentially from one side of the vehicle to the other to capture full⁃cross⁃section imagesof the tunnel. The image stitching software utilizes image matching technology to assemble the full⁃cross⁃section images of the tunnel lining into a panoramic view. The crack detection algorithm employs aninteractive genetic algorithm ( iGA). Using crack images collected from the lining surface of ZhongjiaNo. 2 Tunnel on the Lanzhou⁃Xinjiang Passenger Dedicated Line as an example, the detection resultswere compared and analyzed against those obtained using the harmonic wavelet transform ( HWT)algorithm.