A coherent set of computer vision libraries and extensions to make building experimental CV systems using Python, easy.

Features:

Good integration with numpy and scipy A portable core with few dependencies 3D data processing capabilities integration with openscenegraph

There is no code here yet. I plan to port across my existing own-use code in to this project. Code will be divided up in to a core module and higher level integration modules.

The core module will be targeted at eventual inclusion as a SciKit and will tightly integrate with the imaging SciKit. It will bring together a camera calibration model, 3D geometry model and other basic "building blocks" of 3d computer vision with little dependence on other libraries. It will also include various algorithmic implementations.

The integration modules will then integrate the core modules into higher level libraries for use in visualisation and application building.

Well, that's the plan anyway...