Steps ( Core parameters of the algorithm) – Iterable of iterables (ws, wl, wn, max_iter,). Target_positions ( ( n, 3 ) float) – Target positions assigned to source landmarks Indices and barycentric coordinates ((n,) int, (n, 3) float,). It can also be represented as a tuple of triangle Source_landmarks ( ( n, ) int or ( ( n, ) int, ( n, 3 ) float )) – n landmarks on the the source mesh. It can contain no faces or be a PointCloud. Target_geometry ( Trimesh or PointCloud or ( n, 3 ) float) – Target geometry. Source_mesh ( Trimesh) – Source mesh containing both vertices and faces. Nricp_sumner is less optimized when wn > 0 Nricp_amberg solves for vertex transforms Nricp_sumner solves for triangle positions whereas Nricp_sumner parameters are easier to tune Nricp_sumner tend to preserve more the original shape Only vertices and their neighbors are considered * nricp_amberg fits to the target mesh in less steps The core algorithm is explainedĬomparison between nricp_amberg and nricp_sumner: Nonrigid ICP Algorithms for Surface Registration.”Īllows to register non-rigidly a mesh on another or nricp_amberg ( source_mesh, target_geometry, source_landmarks = None, target_positions = None, steps = None, eps = 0.0001, gamma = 1, distance_threshold = 0.1, return_records = False, use_faces = True, use_vertex_normals = True, neighbors_count = 8 ) Mesh_to_other ( (4, 4) float) – Transform to align mesh to the other objectĬost ( float) – Average squared distance per point Kwargs ( dict) – Passed through to icp, which passes through to procrustes Icp_final ( int) – How many ICP iterations for the closest Icp_first ( int) – How many ICP iterations for the 9 possible Scale ( bool) – Allow scaling in transform Samples ( int) – Number of samples from mesh surface to align Other ( trimesh.Trimesh or ( n, 3 ) float) – Mesh or points in space Mesh ( trimesh.Trimesh object) – Mesh to align with other The principal axes of inertia as a starting point which mesh_other ( mesh, other, samples = 500, scale = False, icp_first = 10, icp_final = 50, ** kwargs ) Īlign a mesh with another mesh or a PointCloud using Transformed ( (n,3) float) – The image of a under the transformationĬost ( float) – The cost of the transformation Matrix ( (4,4) float) – The transformation matrix sending a to b Kwargs ( dict) – Args to pass to procrustes Max_iterations ( int) – Maximum number of iterations Threshold ( float) – Stop when change in cost is less than threshold Initial ( ( 4, 4 ) float) – Initial transformation. Parameters :Ī ( ( n, 3 ) float) – List of points in space.ī ( ( m, 3 ) float or Trimesh) – List of points in space or mesh. Initial transformation can beįound by applying Procrustes’ analysis to a suitable set of landmark Initial transformation is roughly correct. Will only produce reasonable results if the icp ( a, b, initial = None, threshold = 1e-05, max_iterations = 20, ** kwargs ) Īpply the iterative closest point algorithm to align a point cloud withĪnother point cloud or mesh. Functions for registering (aligning) point clouds with meshes.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |