Rotation invariant analysis and orientation estimation method for texture classification based on Radon transform and correlation analysis

Some recent rotation invariant texture analysis approaches such as multiresolution approaches yield high correct classification percentages, but present insufficient noise tolerance. This paper describes a new method for rotation invariant texture analysis. In the proposed method, Radon transform is utilized to project a texture image onto projection space to convert a rotation of the original texture image to a translation of the projection in the angle variable, and then Radon projection correlation distance is introduced. A k-nearest neighbors’ classifier with Radon projection correlation distances is employed to implement texture classification and orientation estimation. Theoretical and experimental results show the high classification accuracy of this approach as a result of using the Radon projection correlation distance instead of repetitious usage of discrete transforms. It is also shown that the proposed method presents high noise tolerance and yields high accuracy in orientation estimation in comparison with Khouzani’s method.