Pseudo-Zernike Moment Invariants to Blur Degradation and Their Use in Image Recognition

Abstract

AbstractThe acquired images often provide a degraded version of the true scene due to the imperfect imaging devices or imaging conditions. Therefore recognition of blurred images has become a key task in pattern recognition and moment invariant-based methods play an important role in this field. In this paper, we construct a new set of invariants using Pseudo-Zernike moments which are invariant to convolution with circularly symmetric point spread function (PSF). The experimental results show that proposed invariants have better performance in terms of invariance and robustness to noise with the comparison to the blur invariants derived from Zernike moments whatever the PSF and noise.KeywordsBlur invariantspseudo-Zernike momentscircularly symmetric blur degradation