Rotation invariant texture recognition using a steerable pyramid

H. Greenspan, S. Belongie, R. Goodman, P. Perona

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review


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, backpropagation and rule-based classifiers. In addition, high accuracy estimation of the input rotation angle is demonstrated.
Original languageUndefined/Unknown
Title of host publicationProceedings of the 12th IAPR International Conference on Pattern Recognition, Vol. 3 - Conference C: Signal Processing (Cat. No.94CH3440-5)
StatePublished - 1 Oct 1994


  • image texture
  • rotation-invariant texture recognition
  • steerable oriented pyramid
  • representative feature extraction
  • input textures
  • filter set
  • invariant representation
  • DFT-encoding step
  • supervised classification
  • k-nearest-neighbours classifiers
  • backpropagation classifiers
  • rule-based classifiers
  • input rotation angle estimation
  • Image databases
  • Feature extraction
  • Spatial databases
  • Discrete Fourier transforms
  • Laplace equations
  • Band pass filters
  • Marine vehicles
  • Image recognition
  • Data mining
  • Degradation

Cite this