Subway lining segment faulting detection based on Kinect sensor

As a kind of tunnel diseases, faulting seriously affect the safety of the tunnel. It is essential to detect lining segment faulting effectively. Compare to traditional detection equipment with high price such as laser camera, a new measurement method for faulting detection by Kinect sensor with low price is proposed. After preprocessing the depth image data, which can be obtained the height difference image by double diagonal difference algorithm. Since the height difference image contains a lot of noise, this paper presents a method of combining the improved median filtering and connectivity domain filtering to solve it. Then recognise faulting, grouting hole and bolt hole through the shape feature. To extract faulting and thinning it. Finally, this paper puts forward a new algorithm called global search, which can identify and calculate the different forms of faulting. The experimental results show that the algorithm can detect and automatically identifies the subway lining segment faulting.