Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
10361766 | Pattern Recognition Letters | 2011 | 9 Pages |
Abstract
⺠A highly discriminative, precise and simple rotation invariant descriptor for natural textures is proposed. ⺠Descriptor is calculated from the statistical pattern of the curvelet coefficients. ⺠A generalized Gaussian density approaches the data distribution for each of the curvelet subbands. ⺠The dominant orientation is estimated in the curvelet space curvelet subbands are reordered according to this reference. ⺠Effectiveness of the proposed descriptor for classification and retrieval tasks.
Keywords
Related Topics
Physical Sciences and Engineering
Computer Science
Computer Vision and Pattern Recognition
Authors
F. Gómez, E. Romero,