some questions about StaMPS step2

[hb z]

Hello,everyone.When I used StaMPS to handle TerraSAR-X data, I encountered some problems in the step 2

The output is shown below
>> stamps(2,2)

########################################
######## StaMPS/MTI Version 3.3 ########
######## Beta version, Sep 2013 ########
########################################

Processing patch directories...
STAMPS:
Processing PATCH /root/Desktop/TSX/INSAR_20160617/PATCH_1
psver currently: 1
psver now set to: 1

########################################
############ StaMPS Step 2 #############
########################################

PS_EST_GAMMA_QUICK: Starting
PS_EST_GAMMA_QUICK: Estimating gamma for candidate pixels
GETPARM: filter_grid_size=50
GETPARM: filter_weighting='P-square'
GETPARM: clap_win=32
GETPARM: clap_low_pass_wavelength=800
GETPARM: clap_alpha=1
GETPARM: clap_beta=0.3
GETPARM: max_topo_err=5
GETPARM: lambda=0.031
GETPARM: gamma_change_convergence=0.005
GETPARM: small_baseline_flag='n'
PS_EST_GAMMA_QUICK: n_trial_wraps=0.103379
PS_EST_GAMMA_QUICK: Initialising random distribution...
PS_EST_GAMMA_QUICK: 577767 PS candidates to process
PS_EST_GAMMA_QUICK: iteration #1
PS_EST_GAMMA_QUICK: Calculating patch phases...
PS_EST_GAMMA_QUICK: Estimating topo error...
PS_EST_GAMMA_QUICK: 100000 PS processed
PS_EST_GAMMA_QUICK: 200000 PS processed
PS_EST_GAMMA_QUICK: 300000 PS processed
PS_EST_GAMMA_QUICK: 400000 PS processed
PS_EST_GAMMA_QUICK: 500000 PS processed
PS_EST_GAMMA_QUICK: gamma_change_change=0.000000
GETPARM: gamma_change_convergence=0.005
PS_EST_GAMMA_QUICK: Finished
STAMPS:
Processing PATCH /root/Desktop/TSX/INSAR_20160617/PATCH_10
psver currently: 1
psver now set to: 1

Only one iteration calculation,I don't know where have problem.Moreover,I have a question for part of the code.


%%% ===============================================
%%% The code below needs to be made sensor specific
%%% ===============================================
if exist(laname,'file')
    la=load(laname);
    inc_mean=mean(la.la)+0.052; % incidence angle approx equals look angle + 3 deg
    clear la
else
    inc_mean=21*pi/180 % guess the incidence angle
end
max_K=max_topo_err/(lambda*rho*sin(inc_mean)/4/pi);

Whether this part need to modify?And how to modify?

[David Bekaert]

One iteration could be ok.

Bit suspicious that the change is gamma_change_change=0.000000.

What happens if you continue for that patch and plot a wrapped ifg for it?

The section of code you show would be ok if you have this la2.mat file.

If the code finds this file it will be using the incidence angles for TSX.

In terms of the range RHO, believe that this one is approximate (average) for a given sensor.

You could take the actual distance here for TSX.

The RHO value is hard-coded at the top of the code.

D.

--
You received this message because you are subscribed to the Google Groups "MAINSAR" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mainsar+unsubscribe@googlegroups.com.
To post to this group, send email to mai...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[hb z]

Hello,David.Thank you for your advice.I continued to step 3,and here is the result:

>> stamps(3,3)

########################################
######## StaMPS/MTI Version 3.3 ########
######## Beta version, Sep 2013 ########
########################################

Processing patch directories...
STAMPS:
Processing PATCH /root/Desktop/TSX/INSAR_20160617/PATCH_1
psver currently: 1
psver now set to: 1

########################################

############ StaMPS Step 3 #############
########################################

PS_SELECT: Starting
PS_SELECT: Selecting stable-phase pixels...
GETPARM: slc_osf=2
GETPARM: clap_alpha=1
GETPARM: clap_beta=0.3
GETPARM: clap_win=32
GETPARM: select_method='DENSITY'
GETPARM: density_rand=20
GETPARM: gamma_stdev_reject=0
GETPARM: small_baseline_flag='n'
GETPARM: drop_ifg_index=
Warning: Not enough random phase pixels to set gamma threshold - using default threshold of 0.3
> In ps_select at 177
  In stamps at 162
PS_SELECT: Initial gamma threshold: 0.300 at D_A=0.03 to 0.300 at D_A=0.40
PS_SELECT: 0 PS selected initially
PS_SELECT: Reestimation gamma threshold: 0.300 at D_A=0.03 to 0.300 at D_A=0.40
PS_SELECT: 0 ps selected after re-estimation of coherence
PS_SELECT: Finished

I think this warning is related to the step 2 that the gamma_change_change=0.000000.
And,what should I do?

在 2017年7月5日星期三 UTC+8下午11:05:50，David Bekaert写道：

[Amir Sharifuddin Ab Latip]

Try to change the filter_grid_size parameter from 50 to 3.

[David Bekaert]

The patch might be empty, do you have water in your area of study?

If so, omit the patch from the patch list and proceed with a next patch.

Cheers,

D.

[hb z]

Hello, Amir. Thank you for your advice, it is very useful. But, when I run step 3, I have some problem. I have 25 patches, most of the patches have the result like this:

>> stamps(3,3)

########################################
######## StaMPS/MTI Version 3.3 ########
######## Beta version, Sep 2013 ########
########################################

Processing patch directories...
STAMPS:
Processing PATCH /root/Desktop/TSX/INSAR_20160617/PATCH_1
psver currently: 1
psver now set to: 1

########################################
############ StaMPS Step 3 #############
########################################

PS_SELECT: Starting
PS_SELECT: Selecting stable-phase pixels...
GETPARM: slc_osf=2
GETPARM: clap_alpha=1
GETPARM: clap_beta=0.3
GETPARM: clap_win=32
GETPARM: select_method='DENSITY'
GETPARM: density_rand=20
GETPARM: gamma_stdev_reject=0
GETPARM: small_baseline_flag='n'
GETPARM: drop_ifg_index=

PS_SELECT: Initial gamma threshold: 1.000 at D_A=0.03 to 1.000 at D_A=0.40

PS_SELECT: 0 PS selected initially

PS_SELECT: Reestimation gamma threshold:  at D_A=0.03 to  at D_A=0.40

PS_SELECT: 0 ps selected after re-estimation of coherence
PS_SELECT: Finished

STAMPS:
Processing PATCH /root/Desktop/TSX/INSAR_20160617/PATCH_10

psver currently: 1
psver now set to: 1

Only 4 of them have PSs selected, like this:

Processing PATCH /root/Desktop/TSX/INSAR_20160617/PATCH_18

psver currently: 1
psver now set to: 1

########################################
############ StaMPS Step 3 #############
########################################

PS_SELECT: Starting
PS_SELECT: Selecting stable-phase pixels...
GETPARM: slc_osf=2
GETPARM: clap_alpha=1
GETPARM: clap_beta=0.3
GETPARM: clap_win=32
GETPARM: select_method='DENSITY'
GETPARM: density_rand=20
GETPARM: gamma_stdev_reject=0
GETPARM: small_baseline_flag='n'
GETPARM: drop_ifg_index=

PS_SELECT: Initial gamma threshold: 0.996 at D_A=0.01 to 1.001 at D_A=0.40
PS_SELECT: 48648 PS selected initially
PS_SELECT: 10000 patches re-estimated
PS_SELECT: 20000 patches re-estimated
PS_SELECT: 30000 patches re-estimated
PS_SELECT: 40000 patches re-estimated
PS_SELECT: 10000 coherences re-estimated
PS_SELECT: 20000 coherences re-estimated
PS_SELECT: 30000 coherences re-estimated
PS_SELECT: 40000 coherences re-estimated
PS_SELECT: Reestimation gamma threshold: 0.996 at D_A=0.01 to 1.000 at D_A=0.40
PS_SELECT: 43120 ps selected after re-estimation of coherence
PS_SELECT: Finished

I don't know what to do. Iam looking forward to your reply.
I am looking forward to yourreply.  
在 2017年7月6日星期四 UTC+8下午4:21:43，Amir Sharifuddin Ab Latip写道：

[hb z]

Thank you for your reply. There is no water in my area of study. And I sloved the problem by changing the filter_grid_size parameter from 50 to 3. But I have new problem in step 3.

I have 25 patches, most of the patches have the result like this:

>> stamps(3,3)

##############################

##########

######## StaMPS/MTI Version 3.3 ########
######## Beta version, Sep 2013 ########
########################################

Processing patch directories...
STAMPS:
Processing PATCH /root/Desktop/TSX/INSAR_20160617/PATCH_1
psver currently: 1
psver now set to: 1

########################################
############ StaMPS Step 3 #############
########################################

PS_SELECT: Starting
PS_SELECT: Selecting stable-phase pixels...
GETPARM: slc_osf=2
GETPARM: clap_alpha=1
GETPARM: clap_beta=0.3
GETPARM: clap_win=32
GETPARM: select_method='DENSITY'
GETPARM: density_rand=20
GETPARM: gamma_stdev_reject=0
GETPARM: small_baseline_flag='n'
GETPARM: drop_ifg_index=

PS_SELECT: Initial gamma threshold: 1.000 at D_A=0.03 to 1.000 at D_A=0.40

PS_SELECT: 0 PS selected initially

PS_SELECT: Reestimation gamma threshold:  at D_A=0.03 to  at D_A=0.40

PS_SELECT: 0 ps selected after re-estimation of coherence
PS_SELECT: Finished

STAMPS:
Processing PATCH /root/Desktop/TSX/INSAR_20160617/PATCH_10

psver currently: 1
psver now set to: 1

Only 4 of them have PSs selected, like this:

Processing PATCH /root/Desktop/TSX/INSAR_20160617/PATCH_18

psver currently: 1
psver now set to: 1

########################################
############ StaMPS Step 3 #############
########################################

PS_SELECT: Starting
PS_SELECT: Selecting stable-phase pixels...
GETPARM: slc_osf=2
GETPARM: clap_alpha=1
GETPARM: clap_beta=0.3
GETPARM: clap_win=32
GETPARM: select_method='DENSITY'
GETPARM: density_rand=20
GETPARM: gamma_stdev_reject=0
GETPARM: small_baseline_flag='n'
GETPARM: drop_ifg_index=

PS_SELECT: Initial gamma threshold: 0.996 at D_A=0.01 to 1.001 at D_A=0.40
PS_SELECT: 48648 PS selected initially
PS_SELECT: 10000 patches re-estimated
PS_SELECT: 20000 patches re-estimated
PS_SELECT: 30000 patches re-estimated
PS_SELECT: 40000 patches re-estimated
PS_SELECT: 10000 coherences re-estimated
PS_SELECT: 20000 coherences re-estimated
PS_SELECT: 30000 coherences re-estimated
PS_SELECT: 40000 coherences re-estimated
PS_SELECT: Reestimation gamma threshold: 0.996 at D_A=0.01 to 1.000 at D_A=0.40

PS_SELECT: 43120 ps selected after re-estimation of coherence
PS_SELECT: Finished

I don't know what to do. Iam looking forward to your reply. Thank you.

在 2017年7月6日星期四 UTC+8下午11:01:50，David Bekaert写道：

[Amir Sharifuddin Ab Latip]

Actually I am not the correct person to ask question. But I suggest to you to analyse your interferograms to see the de-correlated area. Try to reduce the de-correlated area when cropping your images or try to process the small area that contains many buildings/structures.   

--