A multi-scale vector spline method for estimating the fluids motion on satellite images

Abstract : Assessing motion from sequences of atmospheric and oceanographic satellite images is required for environmental simulation, as the dynamic information provided by these images has a strong potential for constraining the models. These images display fluid and potentially turbulent flows. Methods based on vector splines are very efficient to assess motion in this context: they easily allow the conservation law to be taken into account where it is valid and not over the whole image domain; they efficiently implement the 2nd order div-curl regularization, advocated for turbulent fluids, avoiding complex and unstable iterative schemes. The contribution of this paper is to formulate vector splines in a multiscale scheme, with the double objective of assessing motion even in the case of coarse time-sampling and capturing the spectrum of spatial scales associated to turbulent flows. The resulting method only requires the inversion of a band matrix, which is performed by an efficient numerical scheme making the method tractable for large satellite images. Selected results are presented to illustrate the effectiveness of the three characteristics of the approach: the use of control points, 2nd order div-curl regularity, and multiscale formulation.
Complete list of metadatas

Cited literature [15 references]  Display  Hide  Download

https://hal.inria.fr/inria-00264727
Contributor : Jean-Paul Berroir <>
Submitted on : Tuesday, March 25, 2008 - 1:40:08 PM
Last modification on : Friday, May 25, 2018 - 12:02:03 PM
Long-term archiving on : Friday, September 24, 2010 - 11:09:26 AM

Files

rr6478.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : inria-00264727, version 4

Collections

Citation

Till Isambert, Jean-Paul Berroir, Isabelle Herlin. A multi-scale vector spline method for estimating the fluids motion on satellite images. [Research Report] RR-6478, INRIA. 2008, pp.14. ⟨inria-00264727v4⟩

Share

Metrics

Record views

961

Files downloads

834