v3d.PinholeCamera#

class visu3d.PinholeCamera(**kwargs)[source]#

Bases: visu3d.dc_arrays.camera_spec.CameraSpec

Simple camera model.

Camera conventions: In camera/pixel coordinates:

  • u == x == j == w (orientation: →)

  • v == y == i == h (orientation: ↓)

In camera frame coordinates:

  • (0, 0, 1) is at the center of the image

  • x: →

  • y: ↓

  • z is pointing forward

K#

Camera intrinsics parameters.

Type:

FloatArray[’*shape 3 3’]

resolution#

(h, w) resolution

K: FloatArray['*shape 3 3']#
classmethod from_focal(*, resolution: tuple[int, int], focal_in_px: float, xnp: Any | None = None) → visu3d.dc_arrays.camera_spec.PinholeCamera[source]#

Camera factory.

Parameters:

  • resolution – (h, w) resolution in pixel

  • focal_in_px – Focal length in pixel

  • xnp – numpy, jax.numpy or tf.experimental.numpy. Numpy module to use. Default to numpy.

Returns:

A PinholeCamera instance with provided intrinsics.