An Algorithmic Aspect of Zernike Moments in the Presence of Rotation

Zernike moments are widely used in image processing fields (e.g., rotation-invariant watermarking) based on many desirable properties, such as rotational invariance, noise robust and efficient representation of pattern. In image processing operations, an image after rotation will result in a bigger size for keeping the content. This will cause the difference among Zernike moments of an image and its rotated versions. In this paper, we investigate such a problem in detail and introduce a solution by normalizing original and its rotated versions into the same size. Furthermore, we propose an algorithmic scheme of Zernike moments' computation. Both theoretical analysis and experimental testing show that the proposed approach can efficiently avoid the effect of the image size change under rotation.