Sarscape Processing Alos Data practices

1.SARscape InSAR Technology to obtain the Operation flowchart of DEM

2. Data Spotlight

Since the given data is in the ***h1.0_a format (ALOS PALSAR Level1.0, i.e. unprocessed raw signal products, with radiation and geometrical correction parameters. ), so the data focus is processed first.

(SARSACPE's focus module enables fast data focus processing with optimized focusing algorithms for the reflectivity of each point in the raw data, direct output of SLC data, and support for pre-focus raw data mosaicking)

3.DEM Data Format Conversion

Since the downloaded DEM data is in the TIFF format, the Sarscape requires the _dem format, which requires a DEM for format conversion.

In ENVI Classic, choose mosaicking--->georeferenced, the DEM data into the mosaic, and then click Apply to create a tiled dem, the DEM generated here is the standard format of ENVI.

Then select Sarscape/import Data/envi Original to change the output filename suffix *_bil to *_dem. Data Units Select Dem and click Start to execute. You can generate the _dem data required by the Sarscape.

"If there is only one scene TIFF, you can choose Sarscape/import data/generic format/tiff, change the output filename suffix *_bil to *_dem,data units select Dem, click Start to execute"

4. Baseline estimates

Used to evaluate the quality of interference pairs, calculate baselines, orbital offsets (distance to and azimuth), and other system parameters. Interference is only possible when the obtained ground reflection has at least two antennas overlapping, and when the vertical component of the baseline exceeds the threshold, there is no phase information, and the coherence is lost.
Select/sarscape/interferometry/interferometric tools/baseline Estimation

5. Interference graph generation

Select/sarscape/interferometry/phase PROCESSING/1-interferogramgeneration to enter the two-phase SLC data. The Optional Files panel is not set by default.

When processing is complete, the End dialog box pops up and clicks OK. Look at the results under the output path, get the intensity of the master-slave Image: _PWR, interference diagram: _int, after the interference diagram: _dint, synthesis phase: _sint, and oblique distance dem:_srdem. Each result generates a file in TIF format, which is easy to view directly. The results of _dint and _int are as follows.

6. Adaptive filtering and coherence calculation

Interferences generated from SAR data tend to have some noise, and the phase noise caused by ground interference can be removed by means of spatial filtering. At the same time, the coherent graph of interference is generated (describing the phase mass) and the intensity graph of the main image after filtering.
Double-click Sarscape/interferometry/phase PROCESSING/2-Adaptive Filter and Coherence Generation

The result of this step is: Filtered interference diagram: _fint, coherence coefficient diagram: _cc. In the ENVI show the coherence coefficient graph _cc, click to view the cell value, the coherence coefficient distribution at 0-1, the higher the value indicates that the higher the coherence of the region, the smaller the value, the lower the coherence, indicating that the region has changed in two phases.

# #SARscape中干涉图主要滤波-adaptive: This method is suitable for high-resolution data (such as Terrasar-x or cosmo-skymed)

-Boxcar: The frequency of the local interference fringes is used to optimize the filter, which retains the slightest interference fringes as much as possible.

-Goldstein: This filtering method filters are variable, which improves the sharpness of the interference fringes and reduces the incoherent noise caused by the spatial baseline or time baseline. This method is the most commonly used method.

7. Phase unwrapping

The interference phase can only be modeled with 2π, so as long as the phase change exceeds the 2π, it starts and loops again. Phase unwrapping is to solve the problem of 2π ambiguity by phase unwrapping of the phase after smoothing and filtering.
Double-click/sarscape/interferometry/phase PROCESSING/3-Phase unwrapping

# # # Sarscape provides three methods of unwrapping (Unwrapping method Type):

• Regional growth Law (region growing): Choose this method, do not set high coherence threshold (0.15-0.2 is relatively good) in order to leave enough free space for growth, phase mutation part in the entangled image to solve the island existence, this method reduces the error caused by the phase mutation.
• Minimum cost Flow: The default method of unwrapping, when there is a large area of low coherence or other factors limiting the growth of the problem, the least cost flow algorithm can achieve better results than the regional growth method. This method uses square grids, considers all the cells in the image, and masks the pixels with less coherence than the threshold value.
• Delaunay MCF: Unlike the least-cost flow method, this method does not take into account all the pixels on the image, but only the ones that have greater coherence than the threshold value, and instead of using a square grid instead of a droney triangular grid. Only the parts with high coherence are entangled and not affected by the low coherence cells. It is recommended to use this method to minimize the effect of phase mutation in the presence of large numbers of low-coherence objects, such as large amounts of water and dense vegetation in the image.

8. Track refining and re-leveling

when the precise track information is entered, in order to correct the phase shift, the correction of the orbital parameters is done, the corrected result does not generate a new data file, but the information in the wrapped image header file is corrected.

Select/sarscape/interferometry/phase PROCESSING/4-Refinement and reflattening

When using Sarscape for interference processing (INSAR/DINSAR), in a step of the orbital refinement, a control point file (refinement GCP) must be entered to correct the phase of the interference phase and the winding.

Click the button to launch the Control Point selection tool. In the Generate GCP tool in the File selection panel, enter three files, respectively: _upha, _dem, and _fint.

The control points used for track refining generally follow several principles:

• Cannot be located on the residual topography phase of the high frequency
• Can not be located in the deformation phase, away from the deformation area, the control point is considered to be a deformation of 0 points
• Can not be located on the phase transition of the unwrapping error (phase jump)
• If there are regular geometrical errors on the fint, such as track stripes, the GCP is distributed throughout the entire image, thus removing the regular phase error, if it satisfies the above conditions.

Here are a few different scenarios to illustrate the recommendations for control point selection.

If the InSAR is a DEM, select a few control points throughout the image to remove the geometric error (track error) of the system, GCP away from the steep terrain area and the residual topography phase, when the terrain undulating mountains, it is best to choose the bottom of the Valley flat area.

If the dinsar measures the surface deformation, it is possible to select a GCP point from the location of the deformation area without the orbital error.

To do insar measurement of terrain or dinsar measurement of surface deformation, if there is a track error, the need to select multiple GCP points, as a stable reference point, the program from these points to calculate the error phase to remove.

In any case, before selecting GCP, determine the phase where the processing area is not entangled with the error, i.e. phase transitions.

sometimes in order to compare the results, the same region with different observation geometry of the data pair processing, that is, the rising rail data pair and the falling-track data pair. In this case, GCP preferably use the same group, this can reduce the phase error between the two images, the method is: two data pairs of the flat-out after the interference figure _fint geocoding, in the geographical coordinates of the map to select the GCP point.

Finally, there are two things to note: (1) If you want to select a sub-region, it is better to choose a larger area than the study area to ensure that there is a stable region in addition to the deformation area for selecting control points. (2) The orbital refining method has two types: orbital optimization and polynomial optimization. If you select a track optimization method, you need at least 7 control points, otherwise, the polynomial optimization will be automatically changed, the control points are at least the polynomial coefficients (residual Phase Poly degree), otherwise, the polynomial coefficients will be automatically reduced.

9. Phase elevation Conversion--generating DEM

This step is to convert the actual phase, which is completely calibrated and entangled, into a DEM and geocode with the synthetic phase. A DEM file, a coherent image, an image precision image and a resolution image are generated by geocoding.
Select/sarscape/interferometry/phase processing/5a-phase to Height Conversion and Geocoding

10. Phase Transformation and Geocoding

By the absolute calibration and wrapping phase, combined with the synthesis phase, converted to deformation data and geo-coded to the cartographic coordinate system, the default is the LOS direction of the deformation.
Select/sarscape/interferometry/phase processing/5b- Phase to displacement Conversion and Geocoding

Sarscape Processing Alos Data practices