Non-linear Invertible Representation for Joint Statistical and Perceptual Feature Decorrelation
The aim of many image mappings is representing the signal in a basis of decorrelated features. Two fundamental aspects must be taken into account in the basis selection problem: data distribution and the qualitative meaning of the underlying space. The classical PCA techniques reduce the statistical correlation using the data distribution. However, in applications where human vision has to be taken into account, there are perceptual factors that make the feature space uneven, and additional interaction among the dimensions may arise.
In this work a common framework is presented to analyse the perceptual and statistical interactions among the coefficients of any representation. Using a recent non-linear perception model a set of input-dependent features is obtained which simultaneously remove the statistical and perceptual correlations between coefficients. A fast method to invert this representation is also presented, so no input-dependent transform has to be stored. The decorrelating power of the proposed representation suggests that it is a promising alternative to the linear transforms used in image coding, fusion or retrieval applications1.
KeywordsHuman Visual System Joint Statistical Perceptual Feature Perceptual Interaction Image Analysis Application
- 1.K. Fukunaga. Introd. to Statistical Pattern Recognition. Acad. Press, MA, 1990.Google Scholar
- 2.R.J. Clarke. Transform Coding of Images. Academic Press, NY, 1985.Google Scholar
- 6.E.P. Simoncelli and O. Schwartz. Modeling surround suppression in V1 neurons with a statistically-derived normalization model. In M.S. Kearns, editor, Adv. in Neural Inf. Proc. Syst., volume 11. MIT Press, 1999.Google Scholar
- 8.P.C. Teo and D.J. Heeger. Perceptual image distortion. Proc. of the First IEEE Intl. Conf. Im. Proc., 2:982–986, 1994.Google Scholar
- 9.A.B. Watson and J.A. Solomon. A model of visual contrast gain control and pattern masking. JOS A A, 14:2379–2391, 1997.Google Scholar
- 10.I. Epifanio and J. Malo. Linear transform for simultaneous diaginalization of covariance and perceptual metric in image coding. Technical Report 3, Grup de Visió. Dpt. d’Informàtica, Universitat de València, 2000.Google Scholar
- 11.B. Dubrovin and S. Novikov. Modern Geometry. Springer Verlag, NY, 1982.Google Scholar
- 12.G. Legge and J. Foley. Contrast masking in human vision. JOSA, 70:1458–71, 1980.Google Scholar
- 13.B.A. Wandell. Foundations of Vision. Sinauer Assoc. Publish., MA, 1995.Google Scholar
- 14.A.M. Tekalp. Digital Video Processing. Prentice Hall, NJ, 1995.Google Scholar
- 16.D.J. Logan. Non-linear partial differential equations. Wiley&Sons, NY, 1994.Google Scholar
- 17.J. Malo, F. Ferri, R. Navarro, and R. Valerio. Perceptually and statistically decor-related features for image representation: Application to transform coding. Proc. XV Intl. Conf. Patt. Recog., 2000.Google Scholar