Research On Tunnel Crack Image Acquisition And Detection Technology

As a common tunnel disease,tunnel cracks seriously affect the safety of subway operation.Therefore,it is very important to detect cracks on the surface of subway tunnels quickly and efficiently.With the rapid development of computer technology,machine visual technology and image processing technology has gradually become an important method of automatic crack detection.This paper builds a subway tunnel image acquisition system based on machine vision technology,which can achieve fast continuous acquisition.A crack detection algorithm is proposed based on image processing technology to identify the complete cracks and calculate the width.This paper designs the whole section image acquisition scheme for tunnel,elaborates the working principle,and builds a multi-camera and multi-source subway tunnel surface image acquisition system.In order to realize the large area crack detection,provide technical support to 3D reconstruction,this paper studies the image stitching algorithm,and two methods of image stitching method is proposed in this paper,one method is used in ordinary scene,this method uses SIFT algorithm to detect feature points,and then removes the mismatched.A new method of solving transformation model is adopted to make full use of all inner points.Then this method adopts image block projection transform,weighted average fusion in order to remove ghosting,improve quality of image stitching.The other is a fast stitching method under ideal experimental environment,the phase correlation method is used to solve the translation quantity,and the improved optimal joint seam algorithm is used for image stitching.In this paper,image enhancement algorithm is used to balance the illumination of the stitched images and enhance the contrast of crack area and background area.Image segmentation algorithm is used for threshold segmentation and achieving the binary image.Then a filtering algorithm is proposed to remove spot noises,massive noises and line noises.The background information is filtered out,and the binary images only containing crack information is achieved.A method combines binary image and gray image is used to calculate the width of crack.Thinning and removing branches algorithm is used to extract the skeleton of the crack.A normal linear neighborhood of a crack point is picked in the corresponding position of the gray image.A principle called 1-6 is used and the width of cracks is calculated in the pixel domain.The experiment of crack image collection is carried out in the laboratory.In order to verify the stitching effect,images collected on ordinary scene,subway tunnel scene and laboratory scene are tested.Finally,the system acquisition image,long cracked image stitched and the tunnel image with the artificial mark is carried out.The experiment proves that the reliability of the image acquisition system is high and the algorithm of crack detection is more applicable.