Welcome to the long version of the SIO InSAR Cookbook

Note that in order to proceed with this cookbook, you need to have at least a rudimentary knowledge of UNIX commands (e.g. cp, rcp, cd, mkdir, rlogin...) and text-editing (using e.g. vi or textedit). There is a manpage viewer which you can use to determine the correct usage of commands (e.g. test the ones listed in the previous sentence); to utilise it, click here and enter the command name in the Topic field.


(Thanks to Beware of Cat! for the animated cat image.)

Important: Create a working directory named orbit_frame, for example 24707_2925, for each image.

1. Data pre-processing

There are two methods for pre-processing the data. In the first method (steps 1a), the required files (.ldr, .vdf and .raw) are copied from the archive if they are already stored there. If the files are not stored on the archive, you must retrieve them from tape (steps 1b) or cd.

(a) archived data

1.a.1 Locate the .ldr, .vdf and .raw files on the archive.
1.a.2 Copy the files into your working directory.
1.a.3 Continue with steps 1c.

(b) tape retrieval

1.b.1 Locate the appropriate tape and place it into the tape-drive.
1.b.2 Continue with steps 1c.

(c) pre_proc

1.c.1 There are two versions of the pre-processing program, each designed for different data formats (receiving stations): pre_proc_ccrs (Canadian Centre for Remote Sensing) and pre_proc_dpaf (German Processing and Archiving Facility). Usages:
      pre_proc_ccrs   /dev/rmt/0n   orbit_frame
      pre_proc_dpaf   /dev/rmt/0n   orbit_frame
E.g. for a CCRS format image of orbit 24707 frame 2925, type:
      pre_proc_ccrs   /dev/rmt/0n   24707_2925
1.c.2 The pre-processor will write the following files:
      .fix   (.raw file with any missing lines fixed - about 360 Mbytes),
      .PRM   (parameter file - about 1 kbyte), and
      assorted other files (dataheader.log, ldrfile.log, fix.log).
Delete the .ldr, .vdf, .raw and other assorted files. (Do NOT delete the .fix and .PRM files.)
1.c.3 Repeat steps 1a (or 1b) and 1c for each image.

2. SAR processing

The SAR processor reads the .fix file referenced in the .PRM file and writes a .SLC (single-look complex) image file. Since the .SLC file is large (file size = rows x columns x bytes per pixel = 28000 x 6144 x 4 usually = 688128000 bytes = 688 Mbytes), we will simply determine the Doppler center frequency for the first patch of data for each image (a full file is generally processed in ten patches).
2.1 Edit the parameter (.PRM) file for each image; change the number of patches "num_patch" (see
section 1 in the Sample .PRM file) for each to one.
2.2 Run esarp on each .fix file. Make sure to note the Doppler center frequency output by the processor. (Note that the R4 suffix denotes float data format which will create files twice the size of integer two (I2) data type files. It is recommended that you use the I2 option (which is the default) in order to conserve disk space.) Usage:
      esarp   file.PRM   file.SLC   [R4]
E.g. to process the .fix file from step 1.c.1 above, type:
      esarp   24707_2925.PRM   24707_2925.SLC
2.3 Average the Doppler center frequency values.
2.4 Re-edit the .PRM file for each image; change the Doppler center frequency value "fd1" (section 2 in Sample .PRM file) for each to this average value, and change the number of patches back to the correct value (generally ten).
2.5 Repeat step 2.2 (neglecting noting the Doppler center frequency).
2.6 Check that all of the .SLC file sizes are correct (remember that file size = rows x columns x bytes per pixel).

3. Viewing the .SLC file

The .SLC file may be viewed using either ER Mapper (steps 3a) or xgips (steps 3b).

(a) ER Mapper

3.a.1 Run prm2gips to create a GIPS header file from the .PRM file. Note that the R4 option is as for esarp, and if used for esarp should also be used for prm2gips. Usage:
      prm2gips   file.PRM   file.head   [R4]
E.g. to create a GIPS header for 24707_2925.SLC of (default) integer two format, type:
      prm2gips   24707_2925.PRM   24707_2925.head
N.b. .SLC files consist of a real part and an imaginary part, which when using non-I2 format will be represented by two bands in ER Mapper. The amplitude and phase are functions of these real and imaginary parts:
      amplitude:   sqrt(real*real+imaginary*imaginary)
      phase:   atan(imaginary/real)
3.a.2 Utilise GIPS functions to create an amplitude and/or a phase image. For I2 (default) format images, create an amplitude image as outlined in Appendix A, part 1. For non-I2 format, see Appendix A, part 4.
3.a.3 Start ER Mapper. (See Appendix B.)

(b) xgips

3.b.1 Run
prm2gips to create a GIPS header file from the .PRM file. Note that the R4 option is as for esarp, and if used for esarp should also be used for prm2gips. Usage:
      prm2gips   file.PRM   file.head   [R4]
E.g. to create a GIPS header for 24707_2925.SLC of (default) integer two format, type:
      prm2gips   24707_2925.PRM   24707_2925.head
3.b.2 Utilise GIPS functions to create an amplitude and/or a phase image. For non-I2 format images, create an amplitude image as outlined in Appendix A, part 2. For I2 format, see Appendix A, part 3.
3.b.3 Start xgips.

4. Determining image offsets

When determining the offset, you first want to determine a rough offset, and then calculate the precise offset (and stretch parameters). The accuracy required for the rough offset is less than 20 pixels.

(a) determine rough offset

To determine the rough offset you can either use graphical methods (steps 4.a.1 and 4.a.2) or ers_baseline (step 4.a.3).
   offset = repeat coordinates - reference coordinates

4.a.1 Follow steps 3a and Appendix B, part 4 (ER Mapper) or 3b (xgips) for viewing the pair of reference and repeat .SLC images (converted to amplitude).
4.a.2 Determine coordinates for a feature identifiable in both images.
Notes:
(1) if you will be determining the offset using amplitude images in ER Mapper, you will need to multiply the observed y-coordinates by 4 due to the downsampling in Appendix A, part 1;
(2) if you will be determining the offset using xgips, alternately move the mouse and hit the central mousekey to zoom (Figure 4.1) and determine the coordinates of the chosen point (Figure 4.2). Commands to move the screen view etc can be determined by pressing the righthand mousekey and while keeping it depressed, moving downward to highlight Move, and then moving right. Release the mousekey to activate the move.


Figure 4.1: Example zoom window view. Note small cross-hair in the centre.


Figure 4.2: Example pixel coordinates. Note that the relevant coordinates (i.e. those still in the radar frame of reference, and not downsampled) are bracketed: x=673.5, y=838.

4.a.3 Use ers_baseline. Note that this method can be inacurate! Usage:
    ers_baseline   reft0   reftf   ref#   rept0   reptf   rep#
N.b. reft0 is reference image start time, reftf is end time, ref# is satellite (1 = ERS1, 2 = ERS2), rept0 is repeat image start time, reptf is end time and rep# is satellite. Determine the times from the respective .PRM files: start time is "SC_clock_start" (section 4a in Sample .PRM file) and end time is "SC_clock_stop" (section 4b). (The times can be grepped from the .PRM files using the key "_clock_".)
E.g. to run ers_baseline for frame 2925, track 170, orbits 02529 (reference) and 22202 (repeat), type:
    ers_baseline   1995287.7720201157   1995287.7722268077   2   1995286.7719819329   1995286.7721886248   1
N.b. If you calculate a negative yshift, you must add zeros to the front end (top) of the repeat .fix file. To do this, add some error factor to the rough negative offset (e.g. twenty pixels, so a rough offset of -68 pixels would be increased to, say, -88) and then run the add_yshift program. After running add_yshift, the .fix file should be reprocessed using esarp, and the rough offset recalculated (which should then be positive). Usage:
    add_yshift   xnum   yshift   in_file   out_file  
E.g. to add 68 lines of zeros to the .fix file for 22202_2925, type (remembering that since 22202_2925 is an ERS file which have 12060 bytes per line, xnum should be entered as 6030):
    add_yshift   6030   88   22202_2925.fix   22202_2925.fix_new

(b) determine precise offset values and stretch parameters

4.b.1 Edit the .PRM file of the repeat image - enter the rough offset determined in steps 4a above as "xshift" and "yshift" (
section 3a in Sample .PRM file).
4.b.2 Create GIPS headers for the reference and repeat .PRM files using prm2gips. (See step 3.a.1.)
4.b.3 Run offset. Usage:
    offset   reference.head   repeat.head   orbit_offset   [-V].
E.g. to determine offset values for the pair of images 24707_2925 and 5034_2925, type:
    offset   24707_2925.head   5034_2925.head   24707_offset
N.b. The -V option activates the verbose mode (prints the correlations to the screen).
Check the output offset file (orbit_offset) and remove any obvious gross outliers with high SNR (column arrangement is: x xshift y yshift SNR).

(c) fit precise image offset

4.c.1 Run
fitoffset. Usage:
    fitoffset   orbit_offset   [SNR].
Note: SNR implies that you know what cutoff value of SNR you wish to use - any points with SNR less than this cutoff value will be discarded from the adjustment.
E.g. to fit offset determined in steps 4b above which was output to file named 24707_offset type:
    fitoffset   24707_offset
The program will then prompt for a cutoff SNR value - enter a value that you think is appropriate. The program will output the offsets, their slopes and the mean and standard deviation of the residuals based on that cutoff value. The program will then re-prompt for the cutoff SNR value. Continue to enter SNR values until you are happy with the results (located above the data for the individual tie-points). The output consists of six values:
    stretch_r, stretch_a, yshift, sub_int_a, xshift and sub_int_r.
N.b. the decimal part of negative shifts will always be positive, e.g. a shift in the x (range) direction of -2.185326 would be represented as xshift = -3, sub_int_r = 0.814674.
To exit the program, enter some negative SNR value. The program will determine the output based on the last positive SNR value entered.
N.b. it is important to get an accurate value for the offset, as inaccurate values will decorrelate the images (see Figures 4.3 and 4.4).


Figure 4.3: Section of an interferogram.


Figure 4.4: Section of an interferogram. Relatively poor correlation caused by offset difference of 0.52 (2 d.p.) pixels in x and 0.63 (2 d.p.) pixels in y (also slight change in stretch).

4.c.2 Re-edit the .PRM file for each repeat image by pasting the final output from fitoffset. (Edit the shifts (section 3a of the Sample .PRM file) and stretches (section 3b), and add the sub-integer parts of the shifts (section 3c).)
4.c.3 Repeat processing (steps 2) for each of the repeat images.

5. Calculating baseline components

5.1 Run
ers_baseline. Usage is as outlined in step 4.a.3.
5.2 Paste the ers_baseline output into the repeat .PRM file. (See section 5 in Sample .PRM file.)

6. Creating interferometric amplitude and phase images

6.1 Create a GIPS header for each image using
prm2gips.
6.2 Copy header files to orbit_sml.head and edit in order to process only one patch (change "ymax" and "ynum" to 2800 - see section 2 of the Sample GIPS header). E.g. to copy the file created above, type:
      cp 24707_2925.head 24707_sml.head
and then edit the file 24707_sml.head.
6.3 Run phase_grad. Usage:
      phase_grad   reference.head   repeat.head   [topophase.gips]
E.g. to run the program on header files for images 24707_2925 and 5034_2925 type:
      phase_grad   24707_sml.head   5034_sml.head
6.4 Check output amplitude and phase images (amp.gips and phase.gips, respectively) using either ER Mapper (steps 3a) or xgips (steps 3b).
6.5 If the phase image looks OK, continue on to step 6.6. If not, return to the step where you may have gone wrong. Remember when going back past steps in which the .PRM file was edited (i.e. steps 5.2, 4.c.2, 2.4 and 2.1) to remove/replace the added/edited variables.
6.6 Reset your headers to full-size (i.e. change "ymax" and "ynum" back to 28000) and repeat steps 6.3 and 6.4.

7. Removing the topographic signal

7.1 Download
USGS 1:250,000 DEM data from their website.
7.2 Create a basic topographic phase using the DEM data. A sample script for creating the basic topographic phase for the Los Angeles region frame 2925 track 170 is available here. Don't forget to edit the script for your particular data, as outlined in the note at the top of the page.
7.3 Iterate to produce the final (stacked) topographic phase. Sample scripts are available here. Again, don't forget to edit the scripts.
7.4 Create deformation phases using phase_grad with the final topographic phase. Usage:
      phase_grad   reference.head   repeat.head   final_topo_phase.gips
E.g. to run the program on header files for images 24707_2925 and 5034_2925 while using the all_phase_5.gips topo file type:
      phase_grad   24707_2925   5034_2925   all_phase_5.gips
7.5 View the results using either ER Mapper (steps 3a) or xgips (steps 3b).

Sample GIPS header (5034_2925.head)

header = "Image Header"
version = ieee
title = 5034_2925.fix
dtype = a >> 1
dscale = n
dmax = 0.003
dmin = -0.003
dfact = 2.5e-07
dzero = 0
xnum = 6144
xmax = 6144
xmin = 0
ynum = 28000
ymax = 28000
ymin = 0 >> 2
sc_identity = 2
input_file = 5034_2925.SLC
sc_clock_start = 96097.7719396832
sc_clock_stop = 96097.7721325956
local_rad = 6371602.2468
sc_vel = 7125.0383
sc_height = 788084.36
equatorial_rad = 6378137
earth_flat = 0.00335281317789691
near_range = 829924.365777
prf = 1679.902394
rng_samp_rate = 18960000
chirp_slope = 417790000000
pulse_dur = 3.71e-05
lambda = 0.05656
deskew = n
caltone = 0
flip_iq = n
offset_video = n
az_res = 5
nlooks = 1
chirp_ext = 0
scnd_rng_mig = n
rng_spec_wgt = 1
rm_rng_band = 0
rm_az_band = 0
fd1 = 161.887657165
fdd1 = 0
fddd1 = 0
rshift = 4
ashift = 122
sub_int_r = 0.348673
sub_int_a = 0.265091
stretch_r = 0.348673
stretch_a = -6.72228e-05
baseline_start = 97.985
baseline_end = 99.98
alpha_start = 175.364
alpha_end = 175.354
FINIS=

Sample .PRM file (5034_2925.PRM)

num_valid_az = 2800
first_line = 1 >> 6
deskew = n
caltone = 0.000000
st_rng_bin = 1
num_rng_bins = 6144
Flip_iq = n
offset_video = n
az_res = 5
nlooks = 1
chirp_ext = 0
scnd_rng_mig = n
rng_spec_wgt = 1.000000
rm_rng_band = 0.000000
rm_az_band = 0.000000
fd1 = 161.887657165 >> 2
fdd1 = 0.0
fddd1 = 0.0
stretch_r = 2.37174E-04
stretch_a = -6.72228E-05 >> 3b
input_file = 5034_2925.fix
I_mean = 15.436000
Q_mean = 15.496500
rng_samp_rate = 1.896e+07
chirp_slope = 4.1779e+11
pulse_dur = 3.71e-05
radar_wavelength = 0.05656
good_bytes_per_line = 11644
bytes_per_line = 12060
first_sample = 206
SC_clock_start = 96097.7719388426 >> 4a
num_patches = 10 >> 1
near_range = 829924.365777
PRF = 1679.902394
SC_clock_stop = 96097.7721455345 >> 4b
earth_radius = 6371602.2468
SC_height = 788084.36
SC_vel = 7125.0383
yshift = 122
xshift = 4 >> 3a
sc_identity = 2
SC_height = 788084.36
baseline_start = 97.985
baseline_end = 99.980
alpha_start = 175.364
alpha_end = 175.354 >> 5
sub_int_r = 0.348673
sub_int_a = 0.265091 >> 3c

Flowchart of processing steps

Click here to view the flowchart of processing steps.

Flowchart of processing scripts

Click here to view the flowchart of processing scripts.

Appendix A - Brief GIPS tutorial

Note: refer to the GIPS manual or webpages for more detailed instructions and/or utilise man2html to view manpages.
The following are some of the GIPS commands which you are likely to use most frequently:
      git - list image/head values
      iha - image arithmetic
      ihm - merge/resample images
      ihrot - image rotation/reversal/inversion
      ihs - strip image header
GIPS commands are successively performed on images using pipes. That is, the results of one GIPS command are piped to the next command, and so forth. Here are some examples:

1 Create an amplitude image from an I2 format .SLC file for viewing in ER Mapper
Files created using the default I2 format cannot be viewed directly using ER Mapper. In order to create an amplitude image which can be viewed in ER Mapper, first create a GIPS header file for the SLC file (using
prm2gips on the .PRM file). You must then concatenate (cat) that header to the .SLC file, filter the resultant image (using ihconv), resample it (using ihm), invert it E-W (using ihrot, since otherwise you will be viewing the image in the radar frame of reference) and then strip off its GIPS header (using ihs). The complete command string to use is:
      prm2gips   file.PRM   file.head
      cat   file.head   file.SLC   |   ihconv   4   1  /opt/siosar/filters/gauss5x3   |   ihm   -m   "dmin=-1.e-8   dmax=4.e-7   dnull=1.e-5"   |   ihrot   -n   -r   |   ihs   -s   >   file
E.g. to create a viewable amplitude image for 24707_2925.SLC, type:
      prm2gips   24707_2925.PRM   24707_2925.head
      cat   24707_2925.head   24707_2925.SLC   |   ihconv   4   1  /opt/siosar/filters/gauss5x3   |   ihm   -m   "dmin=-1.e-8   dmax=4.e-7   dnull=1.e-5"   |   ihrot   -n   -r   |   ihs   -s   >   24707_2925_amp
N.b. The ihconv arguments used above (4 and 1) force downsampling of the resultant image (by 4 in the y-direction and 1 in the x-direction, respectively). If you wish to use a different sampling rate, adjust the arguments to suit (for another example of downsampling, see Appendix A part 6).

2 Create an amplitude image from an R4 (non-I2) .SLC file for viewing using xgips
To create an amplitude file from an R4 (non-I2) .SLC file, you must create a GIPS header (using
prm2gips on the .PRM file), concatenate (cat) that header to the .SLC file, perform arithmetic (using iha) on the resultant complex image file in order to produce an amplitude image, and then invert that image East-West (using ihrot, since otherwise you will be viewing the image in the radar frame of reference). The complete command string to use is:
      prm2gips   file.PRM   file.head   R4
      cat   file.head   file.SLC   |   iha   -m   "dtype=f"   "p=sqrt(r1*r1+i1*i1)"   |   ihrot   -n   -r   >   file.gips
E.g. to create a viewable amplitude image for 24707_2925.SLC, type:
      prm2gips   24707_2925.PRM   24707_2925.head   R4
      cat   24707_2925.head   24707_2925.SLC   |   iha   -m   "dtype=f"   "p=sqrt(r1*r1+i1*i1)"   |   ihrot   -n   -r   >   24707_2925_amp.gips

3 Create an amplitude image from an I2 format .SLC file for viewing with xgips
As for Appendix A part 1, but without stripping the GIPS header. The complete command string to use is:
      prm2gips   file.PRM   file.head
      cat   file.head   file.SLC   |   ihconv   4   1  /opt/siosar/filters/gauss5x3   |   ihm   -m   "dmin=-1.e-8   dmax=4.e-7   dnull=1.e-5"   |   ihrot   -n   -r   >   file.gips
E.g. to create a viewable amplitude image for 24707_2925.SLC, type:
      prm2gips   24707_2925.PRM   24707_2925.head
      cat   24707_2925.head   24707_2925.SLC   |   ihconv   4   1  /opt/siosar/filters/gauss5x3   |   ihm   -m   "dmin=-1.e-8   dmax=4.e-7   dnull=1.e-5"   |   ihrot   -n   -r   >   24707_2925_amp.gips

4 Create an amplitude image from an R4 (non-I2) .SLC file for viewing in ER Mapper
As for Appendix A part 2, but you must also strip the GIPS header. The complete command string to use is:
      prm2gips   file.PRM   file.head   R4
      cat   file.head   file.SLC   |   iha   -m   "dtype=f"   "p=sqrt(r1*r1+i1*i1)"   |   ihrot   -n   -r   |   ihs   -s   >   file
E.g. to create a viewable amplitude image for 24707_2925.SLC, type:
      prm2gips   24707_2925.PRM   24707_2925.head   R4
      cat   24707_2925.head   24707_2925.SLC   |   iha   -m   "dtype=f"   "p=sqrt(r1*r1+i1*i1)"   |   ihrot   -n   -r   |   ihs   -s   >   24707_2925_amp

5 Create a phase image from an R4 (non-I2) .SLC file
As for the above example, but you must substitute the phase expression for the amplitude expression. The complete command string to use is:
      prm2gips   file.PRM   file.head   R4
      cat   file.head   file.SLC   |   iha   -m   "dtype=f"   "p=atan2(i1/r1)"   |   ihrot   -n   -r   |   ihs   -s   >   file
E.g. to create a viewable phase image for 24707_2925.SLC, type:
      prm2gips   24707_2925.PRM   24707_2925.head   R4
      cat   24707_2925.head   24707_2925.SLC   |   iha   -m   "dtype=f"   "p=atan2(i1/r1)"   |   ihrot   -n   -r   |   ihs   -s   >   24707_2925_phase
Of course, if you wanted a GIPS phase file, you would not strip the header.

6 Downsample a GIPS file
To downsample a GIPS file (i.e. a file which has an incorporated GIPS header file), you must use the
ihm command. So, to downsample a GIPS file originally of size 6144 columns x 28000 rows by a factor of 2 in range and 4 in azimuth, you must type:
      ihm   -m   "xmin=0   xmax=6144   xnum=3072   ymin=0   ymax=28000   ynum=7000"   file.gips   >   out.gips
E.g. to downsample the amplitude GIPS file above, type:
      ihm   -m   "xmin=0   xmax=6144   xnum=3072   ymin=0   ymax=28000   ynum=7000"   24707_2925.gips   >   24707_2925_downsamp.gips
If instead of a GIPS file you were using a stripped file, you would first need to concatenate (cat) the GIPS header. The complete command string to use is:
      cat   file.head   file   |   ihm   -m   "xmin=0   xmax=6144   xnum=3072   ymin=0   ymax=28000   ynum=7000"   >   out.gips
E.g. to downsample the amplitude stripped file above, type:
      cat   24707_2925.head   24707_2925_amp   |   ihm   -m   "xmin=0   xmax=6144   xnum=3072   ymin=0   ymax=28000   ynum=7000"   >   24707_2925_downsamp.gips

7 Subdivide a GIPS file
If instead of downsampling, you wanted to subdivide a GIPS file - say cut out the upper lefthand section - you would use the following command (assuming a file the same size as above, i.e. 6144x28000):
      ihm   -m   "xmin=0   xmax=3072   xnum=3072   ymin=0   ymax=7000   ynum=7000"   file.gips   >   out.gips
E.g. to downsample the amplitude GIPS file for 24707_2925.SLC, type:
      ihm   -m   "xmin=0   xmax=3072   xnum=3072   ymin=0   ymax=7000   ynum=7000"   24707_2925_amp.gips   >   24707_2925_downsamp.gips

8 Create a GIPS header by stripping it from a GIPS file
To create a GIPS header by stripping it from a GIPS file, you must use the
git command. The command string to use is:
      git   -n   file.gips   >   file.head
E.g. to create a header for 24707_2925_amp.gips, type:
      git   -n   24707_2925_amp.gips   >   24707_2925_amp.head

Appendix B - Brief ER Mapper tutorial

Note: refer to the ER Mapper manuals for more detailed instructions.


ER Mapper main menu (standard and common functions). Selected icons, clockwise from top left: 2 Open Algorithm, 3 Copy Window and Algorithm, 9 Go, 11 Stop, 12 Position, 13 zoom, 14 Histogram, 16 Filtering, 17 Equation.

Note: to stop a process at any time, click on the Stop icon of the main menu bar; to start a process, click on the Go icon .

1 Opening an image
1.1 Click on the Open Algorithm icon of the main menu bar - a blank window will appear.
1.2 A selection window will also appear. About three-quarters of the way down the page, under "Files of Type:", the default file type will be listed as Algorithm Data View (.alg). Press the down-arrow to the right and select Raster Image (.ers). Now select your data directory by pressing the topmost down-arrow and working through the directory tree (double-clicking your chosen directories).
1.3 When you have selected your image file, press OK. The image will appear in the window (after a possible delay). Resize the window in the same manner as you would a terminal window (if desired).
Note: To open another window, click on the Copy Window and Algorithm icon . The new window will now become current (designated by asterisks at the top of the window frame). In order to make another window current, move the cursor into the desired window and click the right-hand mousekey.

2 Viewing a non-I2 (R4) amplitude or phase image
2.1 Load the .SLC file as instructed above.
2.2 Click on the Equation icon of the main menu bar.
2.3 The top box (just below the Apply changes button) should read "INPUT1". Edit this to:
      sqrt(input1*input1+input2*input2)
or
      atan(input2/input1)
for the amplitude or phase image, respectively.
2.4 Click on the Apply changes button, then the Close button, and finally on the Go icon of the main menu bar.

3 Filtering the image
3.1 Click on the Filtering icon of the main menu bar.
3.2 Click on the File button and choose Load.
3.3 Select the desired filter (e.g. median 5x5 kernel) and press OK.
3.4 Click on the Close button and then on the Go icon .
Note: In order to remove a filter, click on the Filtering icon, then on the Edit button, choose Delete this filter, then click on the Close button and the Go icon.

4 Zooming and determining image offsets
4.1 Enter zoom mode by clicking on the Zoom icon .
4.2 Move the cursor to the desired image area and then click the left-hand mousekey and drag the zoom box to cover the desired area. Release the mousekey.
Note: To mooz back to the full image, click the right-hand mousekey, move the cursor down to Quick Zoom then to the right and then down to Zoom to Current Dataset.
4.3 Turn of the smoothing (if any) by clicking on the Position icon , clicking on View and selecting Cell Coordinate, then moving the cursor to your selected area and pressing the left-hand mousekey.

Back to: InSAR class homepage
Last modified: 18 October 2000