On image reconstruction, numerical stability, and invariance of orthogonal radial moments and radial harmonic transforms
Chandan Singh, S. Pooja, Rahul Upneja · Pattern Recognition and Image Analysis · 2011
Various radial moments viz. Zernike moments, pseudo Zernike moments, orthogonal Fourier Mellin moments, radial harmonic Fourier moments, Chebyshev-Fourier moments and polar harmonic transforms such as polar complex exponential transforms, polar cosine transforms and polar sine transforms satisfy orthogonal principle. By virtue of which these moments and transforms possess minimum information redundancy and thereby exhibit a good characteristic of image representation. In this paper, a complete comparative analysis is performed by considering image reconstruction capability of each individual moment and transform. The orthogonal properties of above mentioned moments along with the causes of their, reconstruction error, numerical stability and invariance are described.