Abstract
Exponent Fourier moments (EFMs) are suitable for image representation and invariant pattern recognition. EFMs posses more number of uniformly distributed zeros as compared to Zernike Moments. However, these moments tend to be unstable near the center of image and also show a rise in reconstruction error for higher order of moments. In this paper, we propose a new computational framework for calculating the traditional EFM by partitioning the radial and angular part into equally spaced sectors. The proposed approach is simple and results in better image representation capability, numerical stability, and computational speed. Moreover, the proposed approach is completely stable near the center of image.