The role of partitions and functional in descriptor computation for data receptive fields

Abstract

A fundamental concept in Artifitial Vision and Image Processing is that of Complete Description of a data field. From the analitycal point of view completness requires the conservation in the number of degrees of freedom of the visual environment. This constancy has allowed us to establish a special conservation principe, which is determinated by the receptive fields and the functional performed on them.

The purpose of this work is the use of an alternative representation domain, based on the above concepts, and to convergein a spetial non orthonormal transform which combine classic functionlas and data field partitions in order to generate a complete description that, as usual, can be truncated for a particular visual task like a clasification problem.

We have design a clasification system, which is based on a reduced inference system, and we have study the influence of different parameters characteristics of these algebraic analitycal transforms, on the goodness of the clasification system.