Helper routines for using multiple, inexpensive cameras¶

Brandon Jackson and Dennis Evangelista, 2014-2015

Eventually setup cloneable Bitbucket repository:

hg --clone ssh://hg@bitbucket.org/devangel77b/argus

To install the module, run (as administrator):

python setup.py install

This also works although it only gives last stable:

sudo pip install argus

What’s new in Version 0.0.11¶

Hopefully this brings PyPI in line with Bitbucket etc

What’s new in Version 0.0.10¶

Seems to work for Python 3 and OpenCV 3
CMei Omnidirectional Camera stuff
Updated README and documentation
Bug fixes
now available through [https://pypi.org/project/argus/]
API documentation at [http://argus.readthedocs.io/en/latest/]

Description¶

Argus had 1000 eyes. You might only have 3 cameras but hopefully this is useful.

This Python module includes helper routines for calibrating, synchronizing, and using multiple inexpensive cameras for reconstructing 3D positions for scientific uses, such as in biology, biomechanics, ecology, etc. This module includes chessboard calibration, automatic wand tracking, and synchronization using sound and is intended to interface with the python sba sparse bundle adjustment routines for 3D calibration and reconstruction.

Companion Matlab routines are also available.

If these routines are helpful in your work, please cite Jackson, Evangelista, Ray, Hedrick (2016), 3D for the people: multi-camera motion capture in the field with consumer-grade cameras and open source software, Biology Open 5:1334-1342:

@article{Jackson:2016:3D,
author={Jackson, B and Evangelista, D and Ray, D and Hedrick, T},
year={2016},
title={3D for the people: multi-camera motion capture in the field with consumer-grade cameras and open source software},
journal={Biology Open},
volume={5},
pages={1334--1342},
doi={10.1242/bio.018713}}

Still to do¶

Autowand and keypoint finder for automatic detection of extrinsic calibration points. Also provide worked example with sparse bundle adjustment (sba is part of the Python sba library / or is implemented in Matlab in EasyWand).

Thanks to¶

Ty Hedrick, Dylan Ray, Talia Weiss, others