Multiple filterbanks for image processing: implementation issues

Multiple filterbanks have recently been introduced as an effective tool for processing imagedata, in particular for the detection of directional singularities. The practical implementationof such a technique, however, encounters many intrinsic difficulties. Indeed, the fact that different,nontrivial and anisotropic scaling matrices are used at each multiresolution level caneasily result in a highly inefficient way of carrying out the downsampling, upsampling andconvolution operations involved in the process, and of recording and tracking their outputs.The aim of this paper is to propose a general strategy to overcome these implementationissues by properly reformulating the above operations and taking advantage of the representationof the scaling factors in terms of their associated Smith decompositions.