Image processing for multitemporal SAR images

In 1996, the French research group ISIS* proposed a research initiative in the field of radar imaging. One purpose aims to study specificities of multitemporal SAR (synthetic Aperture Radar) images. This paper presents some results of research undertaken in the multitemporal workgroup. In this paper two different kinds of SAR multitemporal data have been distinguished: temporal sequences of spatial images and images containing temporal information like interferograms. To each of them, correspond different types of image processing. The multitemporal information may be used either to enhance static information (multitemporal filtering) or to study temporal evolution (change detection, temporal tracking of structures). In this paper, two different methods of quality restoration are proposed to enhance amplitude data using sequences of SAR images. Then a model for detecting and tracking environmental non-rigid structures (like the coastal line) is explained. Lastly, a segmentation method for phasimetric effects on interferograms is described.

[1]  Donald Geman,et al.  Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images , 1984 .

[2]  Isabelle Herlin,et al.  Cropland detection with SAR interferometry: a segmentation model , 1998, Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181).

[3]  Isabelle Herlin,et al.  A numerical model for large deformation , 1996, Proceedings of 13th International Conference on Pattern Recognition.

[4]  G. Nesti,et al.  Laboratory experiments for the interpretation of phase shift in SAR interferograms , 1997 .