Incremental structure triangulation

class limap.runners.IncrementalStructureTriangulationOptions(max_image_dim: int | None = None, skip_frontend_if_exists_database: bool = False, run_bundle_adjustment: bool = True, use_groups: bool = True, metainfo: MetaInfoComputerOptions = <factory>, _image_description: ImageDescriptionOptions = <factory>, _image_association: ImageAssociationOptions = <factory>, triangulate_points: bool = True, triangulate_lines: bool = True, triangulate_groups: bool = True, point_triangulation: IncrementalTriangulatorOptions = <factory>, line_triangulation: IncrementalLineTriangulatorOptions = <factory>, group_triangulation: IncrementalGroupTriangulatorOptions = <factory>, complete_tracks: bool = True, merge_tracks: bool = True, bundle_adjustment: StructureBundleAdjustmentOptions = <factory>, group_filtering: GroupVerificationOptions = <factory>)

Options for incremental structure triangulation.

__init__(max_image_dim: int | None = None, skip_frontend_if_exists_database: bool = False, run_bundle_adjustment: bool = True, use_groups: bool = True, metainfo: MetaInfoComputerOptions = <factory>, _image_description: ImageDescriptionOptions = <factory>, _image_association: ImageAssociationOptions = <factory>, triangulate_points: bool = True, triangulate_lines: bool = True, triangulate_groups: bool = True, point_triangulation: IncrementalTriangulatorOptions = <factory>, line_triangulation: IncrementalLineTriangulatorOptions = <factory>, group_triangulation: IncrementalGroupTriangulatorOptions = <factory>, complete_tracks: bool = True, merge_tracks: bool = True, bundle_adjustment: StructureBundleAdjustmentOptions = <factory>, group_filtering: GroupVerificationOptions = <factory>) → None

bundle_adjustment: StructureBundleAdjustmentOptions

complete_tracks: bool = True

group_filtering: GroupVerificationOptions

group_triangulation: IncrementalGroupTriangulatorOptions

property image_association: ImageAssociationOptions

property image_description: ImageDescriptionOptions

line_triangulation: IncrementalLineTriangulatorOptions

max_image_dim: int | None = None

merge_tracks: bool = True

metainfo: MetaInfoComputerOptions

point_triangulation: IncrementalTriangulatorOptions

run_bundle_adjustment: bool = True

skip_frontend_if_exists_database: bool = False

triangulate_groups: bool = True

triangulate_lines: bool = True

triangulate_points: bool = True

use_groups: bool = True

limap.runners.incremental_structure_triangulation(options: IncrementalStructureTriangulationOptions, image_dir: Path, model_dir: Path, output_dir: Path, db_path: Path | None = None, structure_db_path: Path | None = None, image_order: list[int] | None = None, neighbors: dict[int, list[int]] | None = None, ranges: tuple[ndarray, ndarray] | None = None) → Path

Incrementally triangulate points, lines, and groups for registered images.

This function assumes images are already registered (poses known). It triangulates the structure (points, lines, groups) incrementally, one image at a time.

Parameters:

options – Configuration with triangulation options
image_dir – Path to images
model_dir – Path to COLMAP model (cameras.bin, images.bin with poses)
output_dir – Output directory for results
db_path – COLMAP database with point correspondences (default: model_dir/database.db)
structure_db_path – Structure database with line/group matches (default: run frontend)
image_order – Order to process images (default: sorted by image_id)
neighbors – Optional image neighbors for frontend
ranges – Optional depth ranges

Returns:

Path to final model directory containing HolisticReconstruction