Nonlinear Operators for Improving Texture Segmentation Based on Features Extracted by Spatial Filtering

M. Unser, M. Eden

IEEE Transactions on Systems, Man, and Cybernetics, vol. 20, no. 4, pp. 804–815, July-August 1990.

Download PDFDownload PSDownload Global BibTeX

An unsupervised texture segmentation system using texture features obtained from a combination of spatial filters and nonlinear operators is described. Local texture features are evaluated in parallel by a succession of four basic operations: (1) a convolution for local structure detection (local linear transform); (2) a first nonlinearity of the form f(x) = |x|^α; (3) an iterative smoothing operator; and (4) a second nonlinearity g(x). The Karhunen-Loève transform is used to reduce the dimensionality of the resulting feature vector, and segmentation is achieved by thresholding or clustering in feature space. The combination of nonlinearities f(x) = |x|^α (in particular, α = 2) and g(x) = log x maximizes texture discrimination, and results in a description with variances approximately constant for all feature components and texture regions. This latter property improves the performance of both feature reduction and clustering algorithms significantly.

@ARTICLE(http://bigwww.epfl.ch/publications/unser9001.html,
AUTHOR="Unser, M. and Eden, M.",
TITLE="Nonlinear Operators for Improving Texture Segmentation Based
	on Features Extracted by Spatial Filtering",
JOURNAL="{IEEE} Transactions on Systems, Man, and Cybernetics",
YEAR="1990",
volume="20",
number="4",
pages="804--815",
month="July-August",
note="")

DOI: 10.1109/21.105080