Title :
Concentric circular sampling for texture analysis
Author :
Arof, Hamzah ; Deravi, Farzin
Author_Institution :
Dept. of Electr. & Electron. Eng., Univ. of Wales, Swansea, UK
Abstract :
This paper introduces new texture descriptors that perform well for rotated and unrotated texture images without the need to be trained with rotated samples. The algorithm uses a circular sampling of the image to produce rotation invariant features. By combining features derived from multiple concentric circles a higher performance in texture classification is achieved
Keywords :
image classification; image sampling; image texture; concentric circular sampling; image texture analysis; multiple concentric circles; rotated texture images; rotation invariant features; texture classification; texture descriptors; unrotated texture images; Algorithm design and analysis; Gabor filters; Image sampling; Image texture analysis; Lattices; Pixel; Sampling methods; Testing; Wavelet analysis; Wavelet transforms;
Conference_Titel :
Image Processing, 1997. Proceedings., International Conference on
Conference_Location :
Santa Barbara, CA
Print_ISBN :
0-8186-8183-7
DOI :
10.1109/ICIP.1997.632050
