Reforming the theory of invariant moments for pattern recognition

摘要

The relationship between the algebraic invariants and the Fourier-Mellin spectrum of an image irradiance distribution is clarified and a new approach for generating the two-dimensional (2D) invariant moments is proposed. Fifty-two invariant functions of the second- to the ninth-order image moments are tabulated, which are invariant with respect to image translation (shift), orientation and size changes and can be used as library functions in computer programs for automatic image identification. The invariant functions derived in this study are shown useful in a numerical example concerned with a character recognition model.