https://github.com/alexflint/triangulation

High quality implementations of many triangulation algorithms in pure python.
https://github.com/alexflint/triangulation

Last synced: 22 days ago
JSON representation

High quality implementations of many triangulation algorithms in pure python.

• Host: GitHub
• URL: https://github.com/alexflint/triangulation
• Owner: alexflint
• Created: 2015-07-02T03:52:36.000Z (over 9 years ago)
• Default Branch: master
• Last Pushed: 2015-09-02T03:12:04.000Z (about 9 years ago)
• Last Synced: 2024-09-24T07:06:59.753Z (about 1 month ago)
• Language: Python
• Homepage:
• Size: 145 KB
• Stars: 33
• Watchers: 4
• Forks: 7
• Open Issues: 1
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Triangulation

[ ![Codeship Status for alexflint/triangulation](https://codeship.com/projects/a56819e0-05cc-0133-6af3-22315beb1c0d/status?branch=master)](https://codeship.com/projects/89535)

High quality implementations of triangulation algorithms in pure python.

### Installation

```bash

pip install triangulation

```

### Usage

```python

from rigidbody import pr, rotation, SE3

from triangulation import triangulate

noise = 1e-3

true_point = np.random.randn(3) + [0, 0, 10]

positions = np.random.randn(num_frames, 3)

orientations = map(rotation.exp, np.random.randn(num_frames, 3)*.1)

poses = [SE3(r, p) for r, p in zip(orientations, positions)]

features = []

for pose in poses:

    z = pr(np.dot(pose.orientation, point - pose.position))

    features.append(z + np.random.randn(2) * noise)

triangulated_point = triangulate(features, poses, algorithm="midpoint")

```

### Algorithms

**`triangulate_midpoint`**

Finds the 3D point minimizing the sum of squared distances to the rays from each camera.

**`triangulate_linear`**

Finds the 3D point minimizing the sum of squared alebraic errors.

**`triangulate_directional`**

Finds the 3D point minimizing the squared directional error in the first view and the view with longest baseline to the first view.

**`triangulate_directional_pair`**

Finds the 3D point minimizing the squared directional error in two views.

**`triangulate_infnorm`**

Finds the 3D point minimizing the infinity norm of the vector of reprojection errors in all views.

**`triangulate_infnorm_fixed`**

Finds any 3D point such that all reprojection errors are no greater than a given threshold.