Forum Replies Created
-
Posts
-
firstly, a general comment: it would be good for you to take a Sarproz course..!!
regarding your questions, a couple of hints:
1. Sarproz by default uses a mask of sparse points based on amplitude local maxima. this helps speeding up a lot the processing, preserving anyway most of the information. also, it is adequate if you want to analyze single strong points. if on the contrary you want to analyze many (all) pixels, you have to force skipping loading that sparse mask or you can create the mask by selecting all pixels (check the corresponding module in the manual).
2. initial and final thresholding of points: lower all thresholds to process/observe as many points as you can. only at the final stage, decide which points to keep, not just based on coherence but also on other parameters
bestcheck in the Sarproz folder. you could find there specific errors caused by the failed connection. if you find them, send them to support. in the meanwhile we sent you a new version by email
you have to update the software using the manage version module
Thank you!
this was indeed a bug introduced by the new windows pointing system.
we will soon update a new version!best
In the image top right corner, beside the title “Reflectivity Map”, you see a list of image tools.
The second from the left, which looks like a comics “speech balloon”, is the “data tips” tool.
Use that tool to select a pixel and consequently run the coherence matrix visualization
those are Matlab features that change format/appearance with time…
bestyou are not obeying to the rules:
https://www.sarproz.com/groups/sarproz-group/forum/topic/forum-general-rules/
https://sarproz.com/manual/Sarproz_errors_management.htmlanyway, yours is an error caused by a new matlab version. we’ll upload a fix soon. in the meanwhile, either change the matlab version to an older one, or switch to the compiled version
best
Hi
the SICD format is not yet supported even if it’s on the todo list
besthi
In the modules APS Estimation, Sparse Points Processing, and Graph Analysis and Refinement, I understand that selecting “Coherence” in the “Weights” option allows SARPROZ to read the interferometric phase from the INTERFEROGRAMS folder and use the magnitude (i.e., coherence) as the weighting factor.
correct
Now, I have developed my own weighting scheme. I would like to ask: can I directly use my weights as the magnitude, construct complex values accordingly, and save them into the interferograms? Does SARPROZ use coherence directly as weights, or does it apply a transformation (e.g., using $\gamma^2 / (1 – \gamma^2)$ instead of $\gamma$)?
yes you can and it will be directly used as weight
It would be very helpful if you could point me to any references regarding the network-based weighted estimation. So far, the only relevant reference I know is:
Perissin and Wang, 2012, Repeat-Pass SAR Interferometry With Partially Coherent Targets, IEEE TGRS.that’s the one!
Hi Marco,
yes, the multi-looking factors helps adjusting the size of pixels with the intent to make them symmetric.
And of course you are right that increasing the size of the filter in one dimension for un-symmetric pixels would make perfect sense. However, after a number of tests, we decided not to implement it, and to keep the size of the filter as small as the smallest resolution allows, based on our experience (or as small as the user likes).
bestHi Marco,
indeed, there is no pre-defined section to explicitly export filtered amplitude images. We planned to implement it, but it was left behind because not much used, this is probably the first time we are asked.
Anyway, you can still do it in a couple of ways.
The first general way is through the View/Export Extended Data/Results.
There, among the parameters, you find at the end of the menu “choose a real file”. you can then navigate to the folder “FILT_AMP” and pick you the image you want. Set manually the Saturation (default for amplitude is 0 3), check B & W as output and there you are. You can plot or export it.
Alternatively, in the multi-channel change detection, you can select “use filtered amp” as input, and then you can plot/export the data. Notice that you can do it for a single multi-channel product with 2/3/4 images or for a sequence.
Hope it is clear enough
bestHi Vitaliano,
the format of the images is not a big problem, you just need to write/read them using a couple of scripts that we can pass you. Some more care should be paid with the meta data, though.
I believe the most effective way to deal with your case is to talk about it in a call.
If good for you, please write an email to admin at sarproz dot com and we try to find a slot in these days
bestThank you Tome
if you send the files to support at sarproz dot com we can create a web link and post it here
bestHi,
yes, of course, you can export SMALL AREA result in Google Earth
you do it from the geocoding module
check here https://sarproz.com/manual/geocod.html if you click on the menu of export options, beside the csv you have also jpegs and various kml formats. kml/kmz are Google Earth formats
bestthe rar file was not uploaded either
13 images for a time series analysis is too limited, use at least 30 images
I suggest you write to support at sarproz dot com to better check the details
then we’ll post here the outcome of the investigation
bestThe log file was not uploaded, try zipping log files in the future
from the inputpar file we see anyway that you are using an old Sarproz version
soon we’ll post here in the forum instructions to download Sarproz compiled with matlab 2022a
try again later with the new version
if issues perists, we’ll start investigating the log file
in the meanwhile, you can check images (searching for issues in the amplitude) and interferograms (to check possible phase jumps).
also, it’s useful to see the images graph to have an idea of your dataset
best