Image analysis by log-polar Exponent-Fourier moments

作者：

Highlights：

• A new set of scaling and rotation-invariant orthogonal moments, named log-Polar Exponent-Fourier moments (LPEFMs), are proposed in log-polar domain.

• A new framework for computing the LPEFMs, which is also suitable for computing existing other harmonic function-based moments, is developed.

• LPEFMs perform much better and faster than conventional methods.

• LPEFMs are validated by image decomposition/reconstruction and robust object recognition.

摘要

•A new set of scaling and rotation-invariant orthogonal moments, named log-Polar Exponent-Fourier moments (LPEFMs), are proposed in log-polar domain.•A new framework for computing the LPEFMs, which is also suitable for computing existing other harmonic function-based moments, is developed.•LPEFMs perform much better and faster than conventional methods.•LPEFMs are validated by image decomposition/reconstruction and robust object recognition.

论文关键词：Exponent-Fourier moments,Log-polar coordinates,Pseudo-polar Fourier transform,Frequency domain interpolation,Scaling and rotation-invariant

论文评审过程：Received 9 October 2018, Revised 31 October 2019, Accepted 15 December 2019, Available online 16 December 2019, Version of Record 11 January 2020.

论文官网地址：https://doi.org/10.1016/j.patcog.2019.107177