Fast computation of Jacobi-Fourier moments for invariant image recognition

•Fast recursive method for the computation of Jacobi-Fourier moments is proposed.•The proposed method not only reduces time complexity but also improves numerical stability of moments.•Better image reconstruction is achieved with lower reconstruction error.•Proposed method is useful for many image processing, pattern recognition and computer vision applications.

The Jacobi-Fourier moments (JFMs) provide a wide class of orthogonal rotation invariant moments (ORIMs) which are useful for many image processing, pattern recognition and computer vision applications. They, however, suffer from high time complexity and numerical instability at high orders of moment. In this paper, a fast method based on the recursive computation of radial kernel function of JFMs is proposed which not only reduces time complexity but also improves their numerical stability.