Clifford Fourier Transform on Vector Fields
Authors:
- 23 May 2005
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Visualization and Computer Graphics
- Vol. 11 (4) , 469-479
Abstract
Image processing and computer vision have robust methods for feature extraction and the computation of derivatives of scalar fields. Furthermore, interpolation and the effects of applying a filter can be analyzed in detail and can be advantages when applying these methods to vector fields to obtain a solid theoretical basis for feature extraction. We recently introduced the Clifford convolution, which is an extension of the classical convolution on scalar fields and provides a unified notation for the convolution of scalar and vector fields. It has attractive geometric properties that allow pattern matching on vector fields. In image processing, the convolution and the Fourier transform operators are closely related by the convolution theorem and, in this paper, we extend the Fourier transform to include general elements of Clifford Algebra, called multivectors, including scalars and vectors. The resulting convolution and derivative theorems are extensions of those for convolution and the Fourier transform on scalar fields. The Clifford Fourier transform allows a frequency analysis of vector fields and the behavior of vector-valued filters. In frequency space, vectors are transformed into general multivectors of the Clifford Algebra. Many basic vector-valued patterns, such as source, sink, saddle points, and potential vortices, can be described by a few multivectors in frequency space.Keywords
This publication has 9 references indexed in Scilit:
- Geometric verification of swirling features in flow fields2003
- Three-dimensional flow characterization using vector pattern matchingIEEE Transactions on Visualization and Computer Graphics, 2003
- Clifford convolution and pattern matching on vector fields2003
- Non-commutative Hypercomplex Fourier Transforms of Multidimensional Signals2001
- Feature extraction of separation and attachment linesIEEE Transactions on Visualization and Computer Graphics, 1999
- Visualizing nonlinear vector field topologyIEEE Transactions on Visualization and Computer Graphics, 1998
- Computationally attractive reconstruction of bandlimited images from irregular samplesIEEE Transactions on Image Processing, 1997
- Signal Processing for Computer Vision1995
- New Foundations for Classical Mechanics1986