An integrated framework for image classification
Xuling Luo, Gagan Mirchandani · 2002
This paper presents a novel method for classifying an image into one of predefined classes in a data bank by applying mutual information to a representation of the Fourier amplitude domain. Template and test images are made translation and rotation invariant through the Fourier-Mellin transform. While mutual information could be employed here, we choose instead to apply it to the lower dimension phase spectrum generated by the complex multiresolution wreath product transform of the Fourier-Mellin amplitude spectrum. The phase information of this transform adequately preserves edges even at lower resolutions while permitting at the same time, a reduction in the computational burden. Brodatz textures and ORL faces are used to demonstrate the capability of this algorithm.