Abstract: As the operation time of subway tunnels increases,common diseases such as lining peeling,cracks,and water leakage are prone to occur due to complex factors such as geological conditions,natural disasters,construction quality,and surrounding environment.At present,the inspection method mainly relies on manual visual inspection,which is limited by various factors such as tunnel section,skylight time,and lighting conditions,making it difficult to improve tunnel inspection with high precision and efficiency.A subway tunnel disease detection system based on a vehicle-mounted multi-eye camera is designed.A special inspection vehicle is used as a carrier platform,and hardware facilities such as linear array cameras,light sources,and power supplies are integrated.The inspection vehicle is used to move along the track at a uniform speed and perform 2D scanning in the vertical direction to collect high-resolution color image data of the tunnel surface.Based on the spliced images,the information of tunnel cracks,leakage,peeling,and other diseases can be judged and identified.Compared with traditional methods,the detection system based on vehicle-mounted linear array cameras can achieve full coverage and rapid collection of tunnel space information,and is convenient for digital archiving and disease development trend analysis.

Key words: linear array camera, tunnel disease, crack detection, image correction