The objectives are as follows: Given an image, it is aimed to obtain an alternative description, from where to proceed to generate a new image, according to certain selection rules, which, in essence, consists in selecting coefficients in a "complete" representation. This is performed in a hierarchycal structure by "computation layers", which generate new spaces (higher leve] tranformated images), the last of which permits a decision layer far visual recognition and/or classification.
© 2008-2024 Fundación Dialnet · Todos los derechos reservados