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

2007/01/20 by Kondor, Risi · 3 citations
#Artificial Intelligence (cs.AI) #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #I.2.10 #I.4.7 #I.5.4

paper · doi:10.48550/arxiv.cs/0701127

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.

Cited by

• Hierarchical Direction Perception via Atomic Dot-Product Operators for Rotation-Invariant Point Clouds Learning
• Kinetics of Peierls dimerization transition: Machine learning force-field approach
• Bispectral OT: Dataset Comparison using Symmetry-Aware Optimal Transport

Related

• Polynomial-based rotation invariant features
• Selective Disk Bispectrum: A Complete and Rotation Invariant Image Descriptor
• The bispectrum as a source of phase-sensitive invariants for Fourier\n descriptors: a group-theoretic approach
• Rotation Differential Invariants of Images Generated by Two Fundamental Differential Operators
• Optimal translational-rotational invariant dictionaries for images