Skip to content
Go to file


FAST-ER and Better: A machine lerning approach to corner detection

Copyright and License

The software is Copyright (c) Edward Rosten and Los Alamos National Laboratory,
2008, Edward Rosten 2017.  There are no restrictions on using this software and
it may only be redistributed under the terms of the GNU General Public License
(a copy of which is included in the file LICENSE). No copyright is claimed on
the output generated by these programs. The files in the fast_trees directory
are generated trivially from the programs included herein, and so are not under

Quick start

Some parts are available without the need to compile or run any programs:

* The best learned FAST-ER tree is in best_faster.tree

* Generated FAST trees for FAST-n and best_faster.tree, along with the trees in
  C++ and MATLAB source code form are in the fast_trees/ directory.

The one you are most likely to want to run is the test_repeatability program. For
that you'll need a dataset:

To get started, you will need a unix or unix-like system (it is likely that
Cygwin and MinGW will work) and to compile and install the following libraries:

* TooN:

* libCVD:

* GVars

Once these have been installed, the project can be compiled using the following

CXX="g++ -std=c++14" ./configure && make

Extensive documentation of the source code and how to run the system is
contained in the HTML reference manual (html/index.html) and the PDF reference
manual (refman.pdf).


FAST-ER and FAST learning code.




No packages published
You can’t perform that action at this time.