Method For Identification Of Rock Discontinuities In Tunnel Face From 3D Point Clouds Based On 3D Laser Scanning

Geological recording is an important work in the tunnel construction,which is of guiding significance to the design and construction of the tunnel.However,the traditional artificial geological sketch is worked in the site of tunnel,it is difficult to quickly and accurately reflect the geological condition of the tunnel.There is an urgent need for a modern and intelligent method for the recognition of tunnel face rock mass.In recent years,some new technology such as photography technology,threedimensional laser scanning technology and so on,was used to record the geological information of the rock mass.These methods are not yet very mature and need to be further studied.According to the possible problems of the current three-dimensional model reconstruction technology of the tunnel,the thesis proposes a projection method based on the spherical surface: the spherical projection twice and then the planar projection method to project the three-dimensional point cloud onto the plane.It can keep the relative topological relationship between the 3D point clouds.And the Delaunay triangulation algorithm is used to reconstruct the three-dimensional surface model of the tunnel to restore the surface characteristics of the tunnel as much as possible.Finally remove the wrong triangular that may exist at the boundary.The thesis is based on a 3D laser scanning point cloud.Combined with the spatial distribution characteristics of the tunnel face and surrounding point clouds,according to the density of the point clouds at the tunnel side wall and the tunnel face,a density clustering method is used to separate the point cloud of the tunnel face from the dataset containing the side wallAccording to the three-dimensional surface model,the occurrence and distance of each triangular are calculated,and the dynamic clustering algorithm is used to cluster the occurrence eigenvalues first,which is used to group the discontinuities,and the distance are clustered to extract a single discontinuity.Then optimize the discontinuities according to the clustering results and extract the feature parameters.The method of finding the minimum cost path is used to identify the structural plane.Combined with the characteristics of three-dimensional spatial data and the cost function in image recognition,a cost function suitable for three-dimensional laser point cloud data is established,and the discontinuity trace is identified successfully.The point clouds of the Ganbagou tunnel is an example of an engineering example,the results show that the method used in this study can separate and reconstruct the face of the tunnel,the rock mass discontinuities and trace on the tunnel face,then the relevant characteristic parameters are calculated successfully.