Invariance analysis of improved Zernike moments

Abstract

Zernike moments have many desirable properties, such as rotation invariance, robustness to noise, expression efficiency, fast computation and multi-level representation for describing the shapes of patterns, but there is a major drawback with Zernike moments: they need to normalize an image to achieve scale invariance. This introduces some errors since it involves the re-sampling and re-quantifying of digital images, and leads to inaccuracy of classifier. In this paper, we present improved Zernike moments, with theory and experiments that show that the improved Zernike moments not only have better rotation invariance, but also have scale invariance. Invariance of the improved Zernike moments shows great improvement over previous methods.