# LICENSE HEADER MANAGED BY add-license-header # # Copyright 2018 Kornia Team # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # """Module for the projection of points in the canonical z=1 plane.""" # inspired by: https://github.com/farm-ng/sophus-rs/blob/main/src/sensor/perspective_camera.rs from __future__ import annotations from typing import Optional import kornia.core as ops from kornia.core import Tensor from kornia.core.check import KORNIA_CHECK_SHAPE def project_points_z1(points_in_camera: Tensor) -> Tensor: r"""Project one or more points from the camera frame into the canonical z=1 plane through perspective division. .. math:: \begin{bmatrix} u \\ v \\ w \end{bmatrix} = \begin{bmatrix} x \\ y \\ z \end{bmatrix} / z .. note:: This function has a precondition that the points are in front of the camera, i.e. z > 0. If this is not the case, the points will be projected to the canonical plane, but the resulting points will be behind the camera and causing numerical issues for z == 0. Args: points_in_camera: Tensor representing the points to project with shape (..., 3). Returns: Tensor representing the projected points with shape (..., 2). Example: >>> points = torch.tensor([1., 2., 3.]) >>> project_points_z1(points) tensor([0.3333, 0.6667]) """ KORNIA_CHECK_SHAPE(points_in_camera, ["*", "3"]) return points_in_camera[..., :2] / points_in_camera[..., 2:3] def unproject_points_z1(points_in_cam_canonical: Tensor, extension: Optional[Tensor] = None) -> Tensor: r"""Unproject one or more points from the canonical z=1 plane into the camera frame. .. math:: \begin{bmatrix} x \\ y \\ z \end{bmatrix} = \begin{bmatrix} u \\ v \end{bmatrix} \cdot w Args: points_in_cam_canonical: Tensor representing the points to unproject with shape (..., 2). extension: Tensor representing the extension (depth) of the points to unproject with shape (..., 1). Returns: Tensor representing the unprojected points with shape (..., 3). Example: >>> points = torch.tensor([1., 2.]) >>> extension = torch.tensor([3.]) >>> unproject_points_z1(points, extension) tensor([3., 6., 3.]) """ KORNIA_CHECK_SHAPE(points_in_cam_canonical, ["*", "2"]) if extension is None: extension = ops.ones( points_in_cam_canonical.shape[:-1] + (1,), device=points_in_cam_canonical.device, dtype=points_in_cam_canonical.dtype, ) # (..., 1) elif extension.shape[0] > 1: extension = extension[..., None] # (..., 1) return ops.concatenate([points_in_cam_canonical * extension, extension], dim=-1) def dx_project_points_z1(points_in_camera: Tensor) -> Tensor: r"""Compute the derivative of the x projection with respect to the x coordinate. Returns point derivative of inverse depth point projection with respect to the x coordinate. .. math:: \frac{\partial \pi}{\partial x} = \begin{bmatrix} \frac{1}{z} & 0 & -\frac{x}{z^2} \\ 0 & \frac{1}{z} & -\frac{y}{z^2} \end{bmatrix} .. note:: This function has a precondition that the points are in front of the camera, i.e. z > 0. If this is not the case, the points will be projected to the canonical plane, but the resulting points will be behind the camera and causing numerical issues for z == 0. Args: points_in_camera: Tensor representing the points to project with shape (..., 3). Returns: Tensor representing the derivative of the x projection with respect to the x coordinate with shape (..., 2, 3). Example: >>> points = torch.tensor([1., 2., 3.]) >>> dx_project_points_z1(points) tensor([[ 0.3333, 0.0000, -0.1111], [ 0.0000, 0.3333, -0.2222]]) """ KORNIA_CHECK_SHAPE(points_in_camera, ["*", "3"]) x = points_in_camera[..., 0] y = points_in_camera[..., 1] z = points_in_camera[..., 2] z_inv = 1.0 / z z_sq = z_inv * z_inv zeros = ops.zeros_like(z_inv) return ops.stack( [ ops.stack([z_inv, zeros, -x * z_sq], dim=-1), ops.stack([zeros, z_inv, -y * z_sq], dim=-1), ], dim=-2, )