A novel set of rotationally and translationally invariant features for images based on the non-commutative bispectrum

TL;DR

This paper introduces a new set of image features that are invariant to rotation and translation, leveraging a generalized bispectrum on the sphere to improve pattern recognition capabilities.

Contribution

It presents a novel approach using non-commutative bispectrum for invariant feature extraction, offering richer image representations than existing methods.

Findings

Features are cubic polynomials in pixel intensities.

Features are invariant to rotation and translation.

Method improves pattern recognition accuracy.

Abstract

We propose a new set of rotationally and translationally invariant features for image or pattern recognition and classification. The new features are cubic polynomials in the pixel intensities and provide a richer representation of the original image than most existing systems of invariants. Our construction is based on the generalization of the concept of bispectrum to the three-dimensional rotation group SO(3), and a projection of the image onto the sphere.