pyfisheye.internal package

Subpackages

• pyfisheye.internal.optimisation package
  • Submodules
  • pyfisheye.internal.optimisation.linear module
    • OptimRuntimeError
    • build_inv_lookup_table()
    • intrinsics_and_z_translation()
    • linear_refinement_extrinsics()
    • linear_refinement_intrinsics()
    • partial_extrinsics()
    • select_best_extrinsic_solution()
  • pyfisheye.internal.optimisation.nonlinear module
    • nonlinear_refinement()
  • pyfisheye.internal.optimisation.optim_result module
    • OptimResult
      • OptimResult.dist_centre
      • OptimResult.extrinsics
      • OptimResult.intrinsics
      • OptimResult.scaling_mat
  • Module contents
    • intrinsics_and_z_translation()
    • linear_refinement_extrinsics()
    • linear_refinement_intrinsics()
    • nonlinear_refinement()
    • partial_extrinsics()
    • select_best_extrinsic_solution()
• pyfisheye.internal.utils package
  • Submodules
  • pyfisheye.internal.utils.check_shapes module
    • check_shapes()
  • pyfisheye.internal.utils.common module
    • build_companion_matrix()
    • compute_image_radius()
    • generate_pattern_world_coords()
    • get_3d_transformation()
    • get_logger()
    • unnormalize_coefficients()
  • Module contents

Submodules

pyfisheye.internal.projection module

pyfisheye.internal.projection.backproject(pixels, intrinsics, distortion_centre, scaling_matrix=None, normalise_rays=True)

Backproject pixels into the world-space.

Parameters:

• pixels (ndarray) – The pixels to backproject. Any shape with last dimension having length 2.

• distortion_centre (ndarray) – The distortion centre x-y in pixels.

• intrinsics (ndarray) – The intrinsic parameters / polynomial coefficients in ascending order.

• scaling_matrix (Optional[ndarray]) – The stretch matrix, leave as None to default to identity.

• normalise_rays (bool) – If set to True, all returned rays will have unit length.

Return type:

ndarray

Returns:

The projected rays with the same shape as pixels except for the last dimension having length 3.

pyfisheye.internal.projection.project(points, intrinsics, distortion_centre, scaling_matrix=None)

Project world points to pixel coordinates. Uses batched method for computing

polynomial roots, but is still quite slow for large numbers of points.

Parameters:

• pixels – The pixels to project (in image coordinates.)

• distortion_centre (ndarray) – The camera’s distortion centre in pixel coordinates.

• scaling_matrix (Optional[ndarray]) – The scaling matrix, leave as None to use identity.

Return type:

ndarray

Returns:

Array with same shape as points except for last dimension having length two for the x-y pixel coordinates.

pyfisheye.internal.projection.project_fast(points, lookup_theta, lookup_rho, distortion_centre, scaling_matrix=None)

Use a lookup table to perform a faster projection of 3D points onto the image space.

Parameters:

• points (ndarray) – The points to project, can be any shape as long as the last dimension has length 2.

• lookup_theta (ndarray) – The theta values in the lookup table in ascending order. Use build_inv_lookup_table() to compute.

• lookup_rho (ndarray) – The rho values in the lookup table. Use build_inv_lookup_table to compute.

• distortion_centre (ndarray) – Distortion centre in pixels.

• scaling_matrix (Optional[ndarray]) – The scaling matrix. Leave as None to use the identity.

Return type:

ndarray

Returns:

The projected pixels.

Module contents