SAR coordinates computation


This topic contains 4 replies, has 2 voices, and was last updated by  periz 2 years, 11 months ago.

  • Author
  • #1202



    I would like to ask if there is any (direct) possibility to compute SAR coordinates (let’s say with the precision of 10-20 pixels) on the basis of known lat, lon. And then, if it is possible to grab out the phase/reflectivity values of (few) selected pixels.

    Thank you, Ivana

  • #1204


    type “fast” in Matlab
    and then use the following expression
    el=1 corresponds to WGS84
    otherwise input a string like GRS80 or the ellipsoid you need

    to extract phase/amp values the easiest way is to use the small area.
    when you load a small area, complex numbers are accessible in the data structure.
    After loading it, type again “fast”
    data are in hand_analisi_rit.WinData
    to get the complex data cube type:

  • #1227


    Hi Daniele,

    thank you, it works. I would like to ask some more:

    I would like to process only 11 points (would appreciate it especially for the sub-cell positioning) but I need to select them. I am afraid that the selecting criteria would be too complicated (various values of intensity etc., scatter plots support > and <, but not =), is it possible to do it manually? I tried (after area selection and the selection of arbitrary 11 points)


    but when opening the “Sparse Point Multi-Image InSAR Analysis”, I can see the original 11 points, not the ones I wanted.

    Does it work somehow this way, or do I need to use some other procedure? In the other variables, I did not find anything interesting. Or what else do I need to set?

    And one more question: I understand the sub-cell positioning can be used for offset-tracking, coregistration check etc. Do I understand well that the sub-cell position in the graphs is in the range direction? Is it also possible (somewhere) to see the interpolated phase and intensity? Is it possible to do the processing with the interpolated values?

    Thank you, Ivana

  • #1258



    could you please answer the questions from the previous one-week-old post in this thread?

    Thank you, Ivana

  • #1262


    right, for some reasons I forgot to answer.
    Well, the easiest way for you to select manually points is to create manually a mask file.
    You can do it in the following way:
    – type fast in the matlab command prompt
    – create vectors of samples and lines of the points you want to analyze:
    Svet=[ … ]; Lvet=[ … ];
    – create a matlab sparse matrix: m=sparse(Svet,Lvet,1,S,L);
    – find a suitable file name (full path pointing to the results folder): fn=[DirResults ‘my_own_mask’];
    – save it: scrivi(fn,m)
    – you should get a message showing you the name of the written file
    now you can either load it in Sarproz by loading a real file, or you can substitute the Sarproz mask file (Maschera.mat in the results folder). In any case, you have to remove/rename Maschera.mat (otherwise, if you load it as a real file, Maschera.mat will be used to read your own mask, and points may not be overlapping). This is achieved e.g. by selecting “all points” in the mask generation module.

    Note 1: the SAR coordinates above are relative to the full site, not to the small area
    Note 2: you can still load your own full site mask in the small area processing

    About subcell positions: in the small area module you can plot the range sub-cell position time series. However, in the full site processing some statistics are calculated both for the range and the azimuth sub-cell positions. At this moment no interpolation of phase/amplitude is carried out. In the next future we will work again on this topic for providing some pixel tracking functions. However, we are not planning functions for interpolating phase/intensities. If you are interested in that, we can plan a project and seek for funds to develop it.

You must be logged in to reply to this topic.