Currently, I’m working with Sentinel-1 GRD data and sometimes the objects likes red rectangled area appears.
Since it have high intensity so object detection detects these noise (ghost?) objects. It degrades desired output.
Therefore, I would like to know what makes these objects and how to handle them, if possible.
The original image is “S1B_IW_GRDH_1SDV_20170829T213232_20170829T213252_007160_00C9E7_2AEE.SAFE”
Many thanks in advance.
+) Plus, it often appears in VH polarization. In the same product, VV polarization does not contain these ghost objects.
Hi, The ‘features’ you have labelled Ghost objects are due to radio frequency interference (RFI) from the ground. Although it can be removed in the processing from raw data (Level 0) to image data (Level 1), this is not performed by the Sentinel-1 Instrument Processing Facility (IPF). It cannot be removed from the processed Level 1 image such as the one you show. RFI can appear in any polarisation but can be more obvious in HV or VH imagery due to the lower radar cross-section of cross-polarisation imagery.
ESA are aware of this issue - it affects < 1% if all S1A/B imagery.
thank you for clarification. I wonder: If they are caused by something on the ground, why are they so regular and exactly perpendicular to the flight direction? Is the effect caused by a single object and then propagated in range direction?
In the S1 raw data, the RFI occur only in a few range lines (probably 7 in this example). There are appox. 1700 range lines per second (the PRF). The azimuth compression part of the SAR processing takes the received energy over the synthetic aperture (approx. 5km in azimuth) to form a single pixel. So if there is a large increase in energy for any of the range lines over the synthetic aperture, the resulting pixel will appear bright. So the azimuth extent of the RFI shows the synthetic aperture length of the SAR.
Within each range line the increase in energy is not continuous but occurs at specific times/ranges. This gives the pattern in range. This pattern depends on the transmitter on the ground and so each RFI looks slightly different.
thanks, @peter.meadows, for the great explanation!
Much thanks, peter.meadows. Your comment help me much!