Forum Replies Created
-
Posts
-
-
1. you did not include all pictures (velocity is not visible). you have to compare all parameters in the same coordinate system (like in sar coordinates)
2. if height and velocity are correlated, something is wrong
3. the range of values to be estimated is too small (both for velocity and height). also, you have lots of images, it means no risk to overfit. take a broader range of values in the estimation
best -
-
you previously posted your question in a discussion on a different topic, so, it has been moved here.
the question you are asking has been discussed in many other posts.
here an example:
https://www.sarproz.com/groups/sarproz-group/forum/topic/slight-offset-after-orthorectification-s1/
please read carefully the forum and the documentation before posting
best -
Hi,
a journal paper that described the core algorithms in Sarproz:
D. Perissin and T. and Wang, “Repeat-Pass SAR Interferometry with Partially Coherent Targets,” IEEE Transactions on Geoscience and Remote Sensing, vol. Volume 50, p. Pages 271–280, 2012.
a conference paper that describes a bit Sarproz:
D. Perissin, Z. Wang, and T. and Wang, “SARPROZ InSAR Tool for Urban Subsidence/manmade Structure Stability Monitoring in China, The,” in Proc. of ISRSE 2011, Sidney (Australia), 2011.
there is then a book chapter that gives a presentation of various InSAR time series processing techniques:
“Interferometric SAR MultiTemporal processing (techniques and applications)” chapter in “Multi Temporal Remote Sensing – Methods and Applications” ISBN 978-3-319-47037-5You can quote any of them if you want.
For what concerns the smart mode algorithm, since it’s a unique feature that Sarproz implements, we decided not to publish any technical detail.
best
-
you are right that a low number of images implies low robustness in the estimation of the coherence and a bias towards high values. That said, numbers are relative and dependent also on other factors. so, it is difficult to tell you what number to use without looking at your dataset and at your results in detail. play with the numbers and observe the consequences.
best -
Hi Paulina,
there are multiple things to consider in your case. An exhaustive answer to your question would require to discuss about various topics. You might want to take a Sarproz course to get more experience (in case, contact courses@sarproz.com). For the moment let me tell you that you should consider the APS estimation as an independent process with respect to the estimation of the parameters of interest. It means, firstly focus on the APS estimation with the points you have, after that, you could try to make a denser estimate in the MISP module (or in the Graph Analysis and Refinement).
best -
Hi Jaroslav,
at the moment, Sarproz saves the 25 reference points in a temporary variable. That variable can be accessed if you run Sarproz in Matlab using the pcodes version. You can access it if you type “fast” in the command prompt when Sarproz is running and the APS module is open and the function for automatic selection of reference points has been executed.
The AutoConnex.mat contains only the results of the graph processing. The reference selection is an operation that comes after graph processing and it is not saved in the AutoConnex.mat. And when you save a reference point, only the chosen reference point is saved. The 25 candidates are only a temporary variable…
best -
the tif file is a 3 channels product. the inputs of the 3 channels (which you can pick up as you like) are all products that can be loaded in the scatter plot module. not as a color clearly (which has to be a scalar), but you indicated you wanted to threshold the MTCD product. so, in the boolean section of the scatter plot module you can implement a selection like e.g. (R > r0 & G < g0) | B > b0 or whatever else you like by playing with <> &| and so on. in this way you select the points you want. then you can either export this selection and use it somewhere else or you can plot your selection in the same module.
again, a sarproz course would be strongly suggested
best -
Hi,
regarding your first question, the outputs of the multi-channel change detection module (shortly described here https://sarproz.com/manual/rgb_module.html) are common outputs that the radar community knows how to interpret. Usually, those outputs are evaluated visually. If you want to set thresholds, you have to build them yourself, e.g. via the “scatter plots” module: https://sarproz.com/manual/xplot.html. If you do not know how to interpret the outputs, I suggest you to take a radar course.
About your second question, yes, it’s just a normalized difference, ad-hoc filtered in space. There are some brief basic tutorials about it on this site (tutorials section), and some brief papers (scientific publications section). Again, probably the most effective way for you to take out the best from these tools is to attend the Sarproz course (a session is starting next week, not sure if seats still available).
best -
if you want to make your own code making your own choices automatically, you are very welcome. you can load the autoconnex.mat in matlab and check all estimated parameters. names should be self explanatory. if you have questions just ask. then you can save the reference point and load it afterwards. you can do it also from the command line.
best -
Sarproz automatically selects the best reference point according to a number of criteria, including those you mentioned.
Nevertheless, in the Sarproz course, we teach how to judge the goodness of the reference point and what to consider to pick up a good one. In most cases, the automatic one is good enough. In some particular situations, the automatic choice might not consider all peculiar details. That’s why some supervision is required.
best -
data are in a folder like : “c:\my project”
you should rename it in something like “c:\myproject” or “c:\my_project” -
Hi,
this is a topic discussed in the course, which I strongly suggest you to take. Generally, better not to estimate the thermal expansion at the APS processing stage. Instead, use a linear model and then estimate the thermal expansion in the MISP module. However, if your area is full of high buildings, (and if you cannot manage picking up points only on the ground) you might be obliged to include also the thermal expansion estimation in the APS module. If this is the case, you have to do this carefully. In any case, you can understand if you are doing things correctly by looking and judging well all outputs that Sarproz offers you during the APS processing. That’s why I suggest you to take the course.
Best -
-
Hi,
please check well the forum before posting questions
you can read all data in Matlab, images and interferograms. Just notice both are complex
Here you can find similar discussions and more links to some explanations about reading data using “leggi”
https://www.sarproz.com/groups/sarproz-group/forum/topic/reading-slc-images-in-matlab/
best
-