PDF Model of Multimedia and Scientific Images in the Wavelet Domain

This work deals with a statistical model of the marginal probability density function (PDF) of digital images in the wavelet domain. Estimation of parameters of the model using least square method is also included. A dyadic decomposition as a special form of the discrete wavelet transform is used in this paper. There is described the statistical model of wavelet coefficients based on the generalized Laplacian. The modeling of PDF in the wavelet domain is well blown procedure, which is used for reconstruction of multimedia data. The multimedia data are characterized by their image content. The robotic astronomical systems, i.e. BOOTES (burst observer optical transient exploring system) provide a huge amount of a data with specific properties (i.e. bit depth, noise, objects). These data are very specific and it is necessary to analyze their PDF in the wavelet domain for their other processing. The parameters of the statistical model of scientific and multimedia data have been estimated and compared. It is very significant for proper setup of the reconstruction algorithms