An RBF Based Method for 2d Image Compression
M.Hossny(1), S. S. EI-Gamal1, H. EI-Shishiny(2) and A. Badr1
(1) Faculty of Computers and Informatics, Cairo University.
(2) IBM Technology Development Center, Cairo, Egypt.
E-mail: ruaab@rusya.eg.net
 

Abstract:
This paper presents another view for image compression. The image is represented as a 3-D graph or surface, where the third dimension corresponds to color z'-ffx.yj. The idea is to find the equation of an implicit surface that realizes z. The results produce comparative quality to the current techniques. However, a great advantage will be gained, that is the fast decoding process. It will be simply achieved by just a matrix multiplication of spatial information by the coefficients of the basis functions.
Keywords: Radial Basis Functions (RBF), Image Compression, Quad-Tree Decomposition.