Applications of mesh parameterization and deformation for unwrapping 3D images of rock tunnels

Panoramic Image Generation Using Centerline- Constrained Mesh Parameterization for Arbitrarily Shaped Tunnel Lining

“…Unwrapping a 3D tunnel mesh as a 2D mesh is a mesh parameterization problem, but if we do not take into account some constraints, the direct unwrapping of the tunnel mesh based on existing algorithms will result in substantial distortions, a loss of details, or the inability to maintain the global shape of the tunnel. Reference [6] analyzed the applicability of two parameterization algorithms (LSCM) [9] and (ABF, ABF++) [10]), and these algorithms exhibited twisted shapes and were not intuitive to interpret. Instead of improving them, the authors [6] proposed a mesh deformation and projection algorithm that is reasonably intuitive to unwrap the 3D mesh as a 2D mesh.…”

“…Reference [6] analyzed the applicability of two parameterization algorithms (LSCM) [9] and (ABF, ABF++) [10]), and these algorithms exhibited twisted shapes and were not intuitive to interpret. Instead of improving them, the authors [6] proposed a mesh deformation and projection algorithm that is reasonably intuitive to unwrap the 3D mesh as a 2D mesh. However, since mesh deformation and projection require many user interactions and expertise, this method results in a low degree of automation and massive distortion.…”

“…The solution is 3D shape estimation, which first uses SFM to recover a sparse 3D point cloud and then uses the point cloud to determine the projection surface for image rectification [7], [8]. However, with an increase in the length of the tunnel, there must be segmented processing; otherwise, error accumulation can cause substantial distortion [6]; thus, the solution is not suitable for constructing a long layout panorama at once. These two-dimensional images are not bidirectional mapping; therefore, any analysis of the image cannot be mapped to the 3D position in the real world.…”

“…Texture reconstruction is performed to map sequenced images to 3D meshes. The 3D meshes have been widely applied to tunnel engineering, such as tunnel inspection [3], building information modeling (BIM) [4], deformation analysis [5], and unwrapping 3D images of rock tunnels [6].…”

“…Therefore, the core problem is to construct a specific energy function to minimize metric distortion. Reference [6] explored two methods to unwrap tunnel mesh: least-squares conformal mapping (LSCM) [10] and angle-based flattening (ABF,ABF++) [11]. These two methods generate twisted shapes and are not intuitive to interpret.…”