A python reimplementation of the C++ image alignment library

Test library now:

Pull code from the repository and run:

$ cd img_alig_lib.python/code
$ python inv_comp.py --video_source='../videos/book1.mp4' --template_corners 106 166 106 69 211 69 211 166 --template_dims 100 100

Incremental Image Alignment Library

This is a library of incremental image alingment algorithms using python 2.7 and OpenCV 2.4.X

Incremental Image alingment is used in tracking applications. In this applications you find the pose of the object in the scene by any means (i.e. object detection) and subsquently the pose is reestimated in the current frame using the pose in the former frame. In order to estimate motion, a model of the displacement of the object is assumed. Therefore, in this context pose stands for the motion parameters.

The elements of any incremental image alignment are:

• The object model (a single texture image, a planar appearance model, AAM, 3D morphable model, etc).
• The motion model (2D similarity, 2D affine, 2D homography, 3D pose (rotation and traslation), etc).
• The optimisation procedure (Gauss-Newton, Levenberg–Marquardt, etc).

Those elements are provided in the library and also an Object Oriented Design that allows to easily introduce efficient technicques for incremental image alignment:

• Hager and Belhumeur's jacobian factorisation method::

  Efficient Region Tracking with Parametric Models of Geometry and Illumination.
  G. Hager, P. Belhumeur
  IEEE Trans. PAMI, 20(10), pp. 1025-39, October 1998.
  

• Baker and Matthews's Inverse Compositional method::

  Lucas-Kanade 20 Years On: A Unifying Framework.
  Simon Baker, Iain Matthews
  International Journal of Computer Vision 56(3), 221–255, 2004
  

The img_aling_lib.python already provides:

• Image Compositional for planar tracking (with a single texture object model) with homography based motion model.
• A demo script for inverse compositional planar tracking.

Requirements

You need the following libraries installed on you system in order to build img_align_lib.python:

• OpenCV <http://www.opencv.org/>
• python 2.7