The bi-orthogonal decomposition in image processing: Signal analysis and texture segmentation

摘要

Decomposition of multi-dimensional signals into orthogonal modes, generated by the data itself, has proven to be a powerful method for the identification of the structures that underlie complex dynamics. This technique is proposed as an image processing tool. The potential applications of the bi-orthogonal decomposition are data compression, noise-resistant communications and texture segmentation. The energy and information content of the bi-orthogonal modes as well as some of the proposed applications are illustrated by several examples.