A toolkit for phase-based image feature detection.
This toolkit consists of a set of functions which use information contained within the phase of a Fourier-transformed image to detect localised features such as edges, blobs and corners. These methods have the key advantage that the properties they measure are invariant with respect to image brightness and contrast.
phasecong- Phase congruency using oriented filtersphasecongmono- Fast phase congruency using monogenic filtersphasesym- Phase symmetry using oriented filtersphasesymmono- Fast phase symmetry using monogenic filters
For more information on a particular function, see the associated docstring and the references therein.
pip install phasepackThese functions were originally written for MATLAB by Peter Kovesi, and were ported to Python by Alistair Muldal. The original MATLAB code, as well as further explanatory information and references are available from Peter Kovesi's website.
MIT License - see LICENSE file for details.