Tom and Hil's website!

C++ source code from my PhD research is now online

License - Download - Browse Source - Pre-requisites - Features - Install - Binaries - Datasets - Parameter documentation - Coming Soon - Known issues - Associated publications

License

License: my code is free for any use. 5-point essential matrix code is for academic use only (based on MATLAB and C code from from Henrik Stewenius), also includes code based on Edward Rosten's FAST code with BSD license. Please email ( ) and let me know if you find it useful or find any bugs or can't get something working.

Download

Works in Windows and Linux. You need svn to get the latest version (or TortoiseSVN in Windows). svn might need to be told if you use a proxy.

BoWSLAM only works in Linux at the moment (wouldn't be too hard to port). Additional instructions here.

Linux: Use this command to check out a copy into a folder called workspace. Hit enter if asked for a password.

svn co https://open.grcnz.com/svn/point-stereo/trunk/workspace workspace

Windows: Right-click in explorer, select SVN Checkout... Enter this url and hit ok:

https://open.grcnz.com/svn/point-stereo/trunk/workspace

Try the latest revision first, revision 955 is fairly stable if anything is broken in the latest one.

Browse Source

https://open.grcnz.com/svn/point-stereo/trunk/workspace/

Or download revision 958 including all code, libraries and binaries (gcc 4.4, some bugs in BaySAC fixed since then)

Pre-requisites:

boost (any recent version, compiled), Eigen (matrix library--version 2.x). Some functions require OpenCV (preferably version 2.0)

Features

(more details here)

BaySAC (and RANSAC and PROSAC and SimSAC) for Essential Matrix estimation and Homography estimation. Topdown refinement of solution. Gold-standard refinement of E (experimental).

Fast 5-point Essential Matrix estimation and Levenberg-Marquardt refinement.

Estimate Homography and decompose into translation, rotation, plane-normal (Levenberg-Marquardt)

Bag-of-Words library: Real-time performance and dynamic retraining. Supports 10k+ images. Fast pairwise correspondences.

K-medoids clustering.

Install

See License_and_Installation_Readme.txt. To use one simple function just copy the relevent code. To build libraries:

Install in Linux (sorry no proper make system yet)

Make three symlinks to the boost, opencv and Eigen in the directory you checked-out into: ln -s PATH_TO_YOUR_OPENCV_DIR opencv; ln -s PATH_TO_YOUR_EIGEN_DIR Eigen; ln -s PATH_TO_YOUR_BOOST_DIR boost

Either:

Open in Eclipse CDT (Galileo). Use Import Existing Project function, or create a new workspace.

The makefiles supplied are just the ones generated from Eclipse but should work. E.g. cd util/Debug; make will build libutil.a with debugging info, etc.

Otherwise can be built with GCC. Consistency checks are added when using -O0.

Install in Windows

Visual Studio 2008 (open ransac_bow_vision_lib.sln)
Visual Studio 2005 (open ransac_bow_vision_lib_VC2005.sln)

Make sure OpenCV, boost, Eigen header and library paths are set in the VC++ Directories

Binaries

The repository is not stable as I keep fixing bugs--it is easiest to get bug fixes using svn.
Download revision 958 including all code, libraries and binaries (gcc 4.4)

You can use the Bag-of-Words code from any programming language under Windows

Datasets

Anyone interested in using datasets should email me ( ). They're mostly too big to put online at the moment but I could post a DVD. Most are low framerate, fairly high resolution and from a single camera, and some have GPS ground truth.

Parameter documentation

Documentation for all parameters is online here. Some headers are documented, e.g. ransac.h for BaySAC, E estimation, H estimation.

Coming soon

1) Proper make system

2) Seperate out clustering functionality (CLARA k-medoids)

3) More examples and documentation :-)

Known issues

K-medoids has hard-coded limit of 450 data-points in Windows, and stack allocation will fail with more than about 1000 in Linux.