Rotation invariant texture analysis using Radon and Fourier transforms

Texture analysis is a basic issue in image processing and computer vision, and how to attain the rotation-invariant texture characterization is a key problem. This paper proposes a rotation-invariant texture analysis technique using Radon and Fourier transforms. This method uses Radon transform to convert rotation to translation, then utilizes Fourier transform and takes the moduli of the Fourier transform of these functions to make the translation invariant. A k-nearest-neighbor rule is employe