homest is a C/C++ library for robust, non-linear homography estimation. A homography (sometimes also called a collineation) is a general plane to plane projective transformation whose estimation from matched image features is often necessary in vision tasks. homest computes a homography estimate that minimizes an appropriate non-linear cost function defined on matching points (currently either non-symmetric transfer error, symmetric transfer error, Sampson error, or reprojection error) and includes robust regression techniques for coping with outliers (i.e., mismatched point pairs). It supports the estimation of fully projective (8 dof) and affine (6 dof) homographies.
|Tags||Scientific/Engineering Image Recognition Libraries|
|Operating Systems||OS Independent|
Release Notes: This release improves thread safety by eliminating global variables. It adds CMake configuration files for building the library.
Release Notes: A couple of issues with memory alignment and random number generation on 64-bit systems were fixed. Extra arguments were added to homography estimation routines which expose the list of detected outliers. Error handling was improved and more verbose output was added.
Release Notes: This release has the ability to estimate affine homographies in addition to fully projective ones. A Matlab MEX interface has been added.