Now days, image compression is essential in transmission and data base storage applications. The current work exposes the design of a compression system using the transform coding technique and in particular using the Wavelet transform. The design is centered on a set of medical images because the high visual quality required for this kind of images. Due to the large diversity of available Wavelet filters, it was chosen a specific filter according to the Best Basis Algorithm. The quantization stage�s design has as support the statistical modeling of the Wavelet coefficients, mainly towards a Laplacian distribution. Following the Linde-Bruno-Gray algorithm is built a general multiresolution scalar quantizer for the whole set of studied images. Finally, the entropy encoding is achieved by means of an arithmetic coder. Inside the document is described the several analysis done to obtain a final design with a compression rate equals to 5.43:1 and preserving the original image quality.

As a tool for analysis the Toolbox of Wavelet of Matlab 5.0 was used. The quantizaton and entropy coding stage's algorithms were created and encapsulated into m-files. Finally a Graphic User Interface (GUI) application in Matlab was created to provide a easy tool for analysis.