Generalized Orthogonal Moment

EasyChair Preprint 4444

Abstract

Orthogonal moments play an important role in image analysis due to their ability to represent images with minimum amount of information redundancy and high level of noise robustness. Recently, several fractional-order orthogonal moments have been proposed. But functions used for the construction of these moments are restricted to fractional-order polynomials. In this paper, orthogonal moments are further generalized to generalized orthogonal moments (GOMs). A general framework is proposed for the construction of functions used in GOMs. Orthogonal polynomials used in traditional orthogonal moments and fractional-order polynomials used in fractional-order orthogonal moments are all special cases of the proposed framework. Properties of the proposed GOMs have been proven. New set of orthogonal moments have also been constructed by choosing several particular functions. Experimental results show the superiority of these moments.

Keyphrases: fractional-order orthogonal moment, generalized orthogonal moment (GOM), image reconstruction, robustness to noise, rotation invariance

@booklet{EasyChair:4444,
  author    = {Yijian Zhang and Zhou Yuan and Jianwei Yang},
  title     = {Generalized Orthogonal Moment},
  howpublished = {EasyChair Preprint 4444},
  year      = {EasyChair, 2020}}