Rotation Invariant Texture Recognition Using a Steerable Pyramid
A rotation-invariant texture recognition system is presented. A steerable oriented pyramid is used to extract representative features for the input textures. The steerability of the filter set allows a shift to an invariant representation via a DFT-encoding step. Supervised classification follows. State-of-the-art recognition results are presented on a 30 texture database with a comparison across the performance of the K-nn, back-propagation and rule-based classifiers. In addition, high accuracy estimation of the input rotation angle is demonstrated
© 1994 IEEE. Date of Current Version: 06 August 2002. This work was supported in part by Pacific Bell, and in part by ARPA and ONR under grant no. N00014-92-J-1860. H. Greenspan was supported in part by an Intel fellowship. P. Perona is supported by NSF Research Initiation grant IRI 9211651 and by an NSF NYI award.
Published - GREcvippr94.pdf