2 Layers Bandelets - Algorithmic Interpretation
Résumé
Nowadays one of the most important issue linked to image transform is to take into account the singularities of a signal which is organized on more than one dimension. The best example is the wavelet transform extension to two dimensional signal analysis. The drawback when one pass from a one dimensional signal process to a two dimensional signal process by simply using separability of wavelet transform is the over representation of irregularities in the wavelet transform domain. In order to decrease this drawback, second generation wavelet transform tries to take geometrical aspects of the image into account in the analysis of the image (one can find examples with bandelets, curvelets, ridgelets and others). 2 layers bandelets or first generation bandelets is among the first wavelet transform which uses the flow to enhance the efficiency of the process. The present proposition is mainly theoretical : we will propose now a practical interpretation of this work in order to make a new implementation of the transform.