Discussion on different interferometric coherence optimization methods
Chemin de la Huniere,
91761 Palaiseau Cedex,
(2) L2S-CNRS, Plateau de Moulon, 91190 Gif-sur-Yvette, France
The interferometric coherence optimisation methods are based on
the definition of so-called projection vectors,
which is the way to combine the information
occurring from the polarimetric channels
of both interferometric images.
An important point is the physical interpretation
of the projection vectors
in terms of scattering charateristics.
An optimisation algorithm has been first proposed by
Papathanassiou, based on the selections of the
best pair of projection vectors. The
Cloude and Papathanassious algorithm leads to
a 3X3 complex
eigenvector problem, each projection vector
representing a selected mechanism for one image.
First we prove that the Cloude and Papathanassiou
algorithm corresponds mathematically to the
most general optimisation of the interferometric
coherence. This means that the interferograms
formed by using these mechanisms have the highest possible
coherence. Nevertheless, a major
drawback of the Cloude and Papathanassiou algorithm is that
in neighbouring resolution cells the optimal
scattering mechanisms can vary significatly,
introducing discontinuities into the extracted
height differences. Moreover, the Cloude and
Papathanassiou algorithm does not fix
uniquely the phase difference between the
two projections vectors.
Although an additional condition can be freely
fixed, an attempt to apply the Cloude and Papathanassiou
algorithm to a selected set of our data
led to an inexploitable height mapping. In this
paper an alternative coherence optimisation algorithm using the
same projection vector for both images is proposed. This method is
tested and compared with the first one in terms of
physical interpretation and results;
the properties of the projection vectors obtained
by both methods are discussed.
Finally, some tests have been been carried out in
order to assess the height estimations
obtained from both methods.