Orthogonal invariant Lagrange-Fourier moments for image recognition

作者：

Highlights：

• A new series of orthogonal moments based on Lagrange polynomials is introduced.

• A new series of feature vectors of images based on orthogonal Lagrange moments is proposed.

• Rotation, Scaling and Translation invariants of the proposed feature vectors are derived.

• Image reconstruction using the proposed orthogonal moments is presented.

• Experimental tests show the efficiency of the proposed feature vectors in pattern recognition and image classification.

摘要

•A new series of orthogonal moments based on Lagrange polynomials is introduced.•A new series of feature vectors of images based on orthogonal Lagrange moments is proposed.•Rotation, Scaling and Translation invariants of the proposed feature vectors are derived.•Image reconstruction using the proposed orthogonal moments is presented.•Experimental tests show the efficiency of the proposed feature vectors in pattern recognition and image classification.

论文关键词：Lagrange polynomiales,Orthogonal Lagrange-Fourier moments,Multi-channel Lagrange-Fourier moments,Quaternion Lagrange-Fourier moments,Image recognition

论文评审过程：Received 29 August 2021, Revised 19 January 2022, Accepted 28 March 2022, Available online 8 April 2022, Version of Record 21 April 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117126