2009

During the first term of 2009, an experimental topic dealing with corner reflectors has been developed. Based on a wood structured three parts kit so that a man can carries it inside a rucksack, and easy to build on site (near a glacier or on a glarcier for example), they allow either a good positionning of a given landmark, or a measure of the displacement.

After a first test near Lognan seracs, a twin system has ben put in the middle part of glacier d'Argentière. An other has been located near Luc Moreau's camera (right bnak of Glacier d'Argentière).

2008

In order to select CS candidates, both automatic tools and interactive tools have to be developped. Yet, a main topic has been emphasized : graphic tools to interpret temporal stacks of SAR images.

To cope with this topic, some fundamental analysis about resampling has  to be undertaken. As SAR images are complex data (a pixel has a real part and an imaginary part), a specific analysis of 2-D complex data has been investigated. Actually, there is no global work about this domain (complex 2-D data) and we have found some elementary knowledges in various domains as Solid State physics (the theory of dislocations), mathematics (phase vortex), ultrasoud echography (correlated speckle),... It seems there is a basic clue in the analysis of phase vortex in a 2-D space. Due to these phase vortices, an oversampled SAR image has a very characteristic trend : indeed, a phase vortex corespond to a null amplitude value. Even if a SAR image has no null value, the corresponding oversampled image has a lot of very low level values based on specific phase vortices.

It is why a specific research work devoted to phase vortex has been undertaken in order to develop MAGNICOS tool. During 2008, we have define a specific milestone about this topic, yielding a proposition to GRETSI 2009 (Dijon). A research report will be provided about phase vortex in july 2009 as a non scheduled deliverable.


At the same time, a beta test version of MAGNICOS has been provided as a contractual deliverable. Yet, some troubles have been pointed on a graphical software used by this tool : XView, well known in the Linux community. Indeed, XView does not exist for 64 bits computers. It is why we have planned for 2009 an update of XView software to cope with this problem.