Abstract
While clamping the fractal operator to the grayscale range of values of the images preserves its contractivity, performing the rounding process to discrete levels spoils this property and the given iterative sequence is generally not convergent. In practice this lack of convergence is not observed by HVS (Human Visual System) on decoder’s output. We explain this phenomenon by presenting a strict notion of the perceptual convergence at the given threshold. It is used to prove that any iterative sequence for a fractal operator which is contractive in l ∞ norm with contractivity c *<1, after clamping and rounding to integer levels, is perceptually convergent at the threshold τ≥1/(1−c *).
Preview
Unable to display preview. Download preview PDF.
References
Barnsley, M.F.: Fractals everywhere. Addison Wesley, Reading, MA (1988)
Barnsley, M.F., Hurd, L.P.: Fractal image compression. AK Peters. Ltd, Wellesley, MA (1993)
Dugundi, J., Granas, A.: Fixed point theory. Polish Scientific Publishers, Warszawa (1982)
Fisher, Y. ed.: Fractal image compression, Theory and Applications. Springer Verlag, New York (1995)
Jacquin, A.E.: Image coding based on a fractal theory of iterated contractive image transformations. IEEE Trans. on Image Processing 1 (1992) 18–30
Skarbek, W.: On convergence of affine fractal operators. Image Processing and Communications 1 (1995) 33–41
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1999 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Skarbek, W. (1999). Perceptual convergence of discrete clamped fractal operator. In: Raś, Z.W., Skowron, A. (eds) Foundations of Intelligent Systems. ISMIS 1999. Lecture Notes in Computer Science, vol 1609. Springer, Berlin, Heidelberg . https://doi.org/10.1007/BFb0095126
Download citation
DOI: https://doi.org/10.1007/BFb0095126
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-65965-5
Online ISBN: 978-3-540-48828-6
eBook Packages: Springer Book Archive