Texture Characterization via Joint Statistics of Wavelet Coefficient Magnitudes

Eero P Simoncelli and Javier Portilla

Published in:
Proc. 5th Int'l Conference on Image Processing
Chicago, IL. October 4-7, 1998.
doi: 10.1109/ICIP.1998.723417
© IEEE Signal Processing Society.

Presentation slides (pdf, 800k).

Full-length journal article: Int'l Journal of Computer Vision, 2000


We present a parametric statistical characterization of texture images in the context of an overcomplete complex wavelet frame. The characterization consists of the local autocorrelation of the coefficients in each subband, the local autocorrelation of the cofficent magnitudes, and the cross-correlation of coefficient magnitudes at all orientations and adjacent spatial scales. We develop an efficient algorithm for sampling from an implicit probability density conforming to these statistics, and demonstrate its effectiveness in synthesizing artificial and natural texture images.
Full Text (328k, gzip'ed PostScript) / Full Text (284k, pdf) / Online Description & Examples / EPS Online Publications