PROJECT TITLE :

Template-Free Wavelet-Based Detection of Local Symmetries - 2015

ABSTRACT:

Our goal is to detect and cluster totally different kinds of native symmetries in pictures in an exceedingly scale- and rotation-invariant way. We have a tendency to propose an economical wavelet-based mostly method to work out the order of native symmetry at every location. Our algorithm relies on circular harmonic wavelets which are used to get steerable wavelet channels corresponding to different symmetry orders. To offer a live of native symmetry, we have a tendency to use the F-take a look at to examine the distribution of the energy across different channels. We have a tendency to provide experimental results on artificial pictures, biological micrographs, and electron-microscopy images to demonstrate the performance of the algorithm.