Fourier Analysis of Orientation Histograms for Texture Classification

M. Alemán-Flores and L. Álvarez-León (Spain)


Image Processing and Analysis, Texture Analysis, Edge Orientation, Fourier Transform


This work presents an approach to texture classification in which orientation histograms have been generated to compare textured regions. The orientation and magnitude of the gradient in every point of a texture are estimated and, combining them, an orientation histogram is built for each texture. Fourier analysis is used to measure the similarity of the histograms, considering the effects of a change in the size or orientation of the image. We have introduced a weighting function to reduce the influence of noise. Furthermore, we have tested the robustness of our method when some grayscale transformations are performed on the images.

Important Links:

Go Back