Omnidirectional camera calibration in python

Last update: Nov 22, 2022

Related tags

Deep Learning py-omnicalib

Overview

Omnidirectional Camera Calibration

Key features

pure python
initial solution based on A Toolbox for Easily Calibrating Omnidirectional Cameras (Davide Scaramuzza, Agostino Martinelli, Roland Siegwart)
Optimized solution via Levenberg–Marquardt algorithm on reprojection error
Lie algebra based extrinsics optimization

(undistort example)

Usage

Corner detection (see python -m omnicalib.detect --help for detailed argument description)

python -m omnical.detect --chessboard <rows> <columns> <square-size> --max-dim <max-dim> --threads <threads> <image-folder>

The file detections.pickle is written, which contains a pickled dictionary with the following format

{
    'detections': {
        <image_path_0>: {'image_points': <image_points>, 'object_points': <object_points>},
        ...
}

Image paths (pathlib.Path) are absolute
image_points are torch.Tensor with shape N x 2 and dtype torch.float64
object_points are torch.Tensor with shape N x 3 and dtype torch.float64 and third colum, the z coordinate, all zeros

If an external corner detection method is used, this file can simply be written manually.

Calibration (see python -m omnicalib --help for detailed argument description)

python -m omnical --degree <degree> detections.pickle

Result

A calibration.yml file (see example_calibration.yml) containing

extrinsics as list of 3 x 4 matrices
poly_incident_angle_to_radius, a polynom that converts incident angle to image radius
poly_radius_to_z, a polynom that converts radius to z component of view vector
principal_point in pixel

Method

For a detailed description see method.md.

Simulation

Play with the calibration on simulated data with the provided jupyter lab notebook. It generates randomized data for the following ideal projection models

equidistant
stereographic
orthographic
equisolid

, see Large Area 3D Human Pose Detection Via Stereo Reconstruction in Panoramic Cameras for a detailed description.

Example

Jupyter lab notebook, example results

You might also like...

Improving Calibration for Long-Tailed Recognition (CVPR2021)

19 Apr 28, 2021

Pytorch Implementation of Spiking Neural Networks Calibration, ICML 2021

SNN_Calibration Pytorch Implementation of Spiking Neural Networks Calibration, ICML 2021 Feature Comparison of SNN calibration: Features SNN Direct Tr

60 Dec 27, 2022

S2-BNN: Bridging the Gap Between Self-Supervised Real and 1-bit Neural Networks via Guided Distribution Calibration (CVPR 2021)

S2-BNN (Self-supervised Binary Neural Networks Using Distillation Loss) This is the official pytorch implementation of our paper: "S2-BNN: Bridging th

52 Dec 24, 2022

[CVPR 2022] Official code for the paper: "A Stitch in Time Saves Nine: A Train-Time Regularizing Loss for Improved Neural Network Calibration"

MDCA Calibration This is the official PyTorch implementation for the paper: "A Stitch in Time Saves Nine: A Train-Time Regularizing Loss for Improved

21 Dec 22, 2022

Omnidirectional camera calibration in python

Related tags

Overview

Omnidirectional Camera Calibration

Key features

Usage

Result

Method

Simulation

Example

You might also like...

Improving Calibration for Long-Tailed Recognition (CVPR2021)

Pytorch Implementation of Spiking Neural Networks Calibration, ICML 2021

S2-BNN: Bridging the Gap Between Self-Supervised Real and 1-bit Neural Networks via Guided Distribution Calibration (CVPR 2021)

the code for paper "Energy-Based Open-World Uncertainty Modeling for Confidence Calibration"

We utilize deep reinforcement learning to obtain favorable trajectories for visual-inertial system calibration.

Improving Calibration for Long-Tailed Recognition (CVPR2021)

Source code of NeurIPS 2021 Paper ''Be Confident! Towards Trustworthy Graph Neural Networks via Confidence Calibration''

CSAC - Collaborative Semantic Aggregation and Calibration for Separated Domain Generalization

[CVPR 2022] Official code for the paper: "A Stitch in Time Saves Nine: A Train-Time Regularizing Loss for Improved Neural Network Calibration"

Releases(1.0.1)

1.0.1(Mar 1, 2022)

1.0.0(Dec 21, 2021)

Owner

Thomas Pönitz

Vehicle Detection Using Deep Learning and YOLO Algorithm

Referring Video Object Segmentation

Code for Parameter Prediction for Unseen Deep Architectures (NeurIPS 2021)

mbrl-lib is a toolbox for facilitating development of Model-Based Reinforcement Learning algorithms.

Official Pytorch implementation of Meta Internal Learning

Image-retrieval-baseline - MUGE Multimodal Retrieval Baseline

Research code for the paper "How Good is Your Tokenizer? On the Monolingual Performance of Multilingual Language Models"

Pytorch implementation of "M-LSD: Towards Light-weight and Real-time Line Segment Detection"

An implementation of MobileFormer

Code for the SIGGRAPH 2022 paper "DeltaConv: Anisotropic Operators for Geometric Deep Learning on Point Clouds."

3rd Place Solution of the Traffic4Cast Core Challenge @ NeurIPS 2021

Code examples and benchmarks from the paper "Understanding Entropy Coding With Asymmetric Numeral Systems (ANS): a Statistician's Perspective"

An implementation of the Contrast Predictive Coding (CPC) method to train audio features in an unsupervised fashion.

Toontown House CT Edition

Pytorch Implementation for (STANet+ and STANet)

Open source code for the paper of Neural Sparse Voxel Fields.

Numerai tournament example scripts using NN and optuna

BasicNeuralNetwork - This project looks over the basic structure of a neural network and how machine learning training algorithms work

Markov Attention Models

Mall-Customers-Segmentation - Customer Segmentation Using K-Means Clustering