Deformation Measurement Of Tunnel Surrounding Rock Based On Digital Image Processing

With the rapid growth of our national economy,the construction of transportation network has been developing gradually.Because the tunnel has the advantages of small environmental impact and can shorten the distance of transportation,it becomes an important part of urban subway,expressway and high-speed railway.Considering that the tunnel is prone to deformation during construction and operation,which leads to the occurrence of a safe disaster,real-time monitoring of tunnel deformation becomes an indispensable task.However,the existing methods of tunnel deformation monitoring are vulnerable to the constraints of on-site conditions and large disturbance to construction.This paper studies a tunnel deformation monitoring system based on digital image processing.The system has the advantages of small influence on construction,high degree of automation,low cost and strong maneuverability.The basic realization process is described as follows.Firstly,the laser transmitter is fixed to the measured point of the surrounding rock.By adjusting the "well" character target,the laser beam emitted by the laser generator just happens to form a clear facula on the target.Then the digital camera is facing the target of "well" to obtain the image data of the facula on the "well" target.Because the "well" target is composed of the blue and red grid lines with precise scale(see Figure 2-2),each blue mesh is 25 mm×25 mm and each red mesh is 5 mm×5 mm.The pixel coordinates of the laser spot can be extracted by accurately locating the specific color area of the facula in the "well" target.Finally,the real spatial location results of the spot are obtained through the transformation of the object coordinates and the pixel coordinates.The displacement of the light facula generated by the laser emitter fixed on the surrounding rock is the deformation displacement of the surrounding rock.The image processing algorithms used in this system include:(1)Template matching method is adopted in the system.The template matching method is used to extract the cross points in the blue and red grid images after color segmentation,which lays the foundation for the reconstruction of the grid space.(2)Connected component labeling is adopted in the system.The connected component labeling algorithm is used to mark the intersection points of the blue grid image,and the projection method is used to complete the reconstruction of the spatial position of the grid so as to facilitate the regional positioning and fine positioning of the subsequent facula.(3)Threshold segmentation and centroid method is adopted in the system.The laser point is captured in the image as a small area.Therefore,in order to accurately extract the central pixel position of the laser,the laser area is extracted by threshold segmentation firstly.Secondly,the center of the facula is extracted by the centroid method.Finally,the spot coordinates are determined by the Euclidean distance between the central point of the facula and the blue cross point.(4)Nearest neighbor method is adopted in the system.The nearest neighbor method is used to determine the coordinates of the closest red cross point to the facula,and the accurate location of the facula image in the red grid image is obtained.Moreover,the final location result of the facula is obtained by the space coordinate transformation.Through monitoring experiments,it is verified that the method of digital image processing to monitor the deformation of tunnel surrounding rock can reach 0.108 mm accuracy under certain conditions.The experimental results show that the deformation monitoring algorithm studied in this paper has some robustness and meets the accuracy requirements of ± 0.5 mm for vertical displacement and ± 3 mm for horizontal displacement in the "Code for Urban Mass Transit Engineering Survey"(GB/T 50308—2017).