Hello Step Forum!
I am currently taking my master degree in Remote Sensing for Archaeology and I am new to this field and have many questions.
One such question is about TerraSAR-X image products - MGD (multilook ground range detected):
As far as I understand, multi looking is the process of joining multiple single look images or multiple bands of a same image to remove speckle, at the cost of resolution, is this correct?
If so, how is ground range corrected image of TSX (which is the normalization of pixel size - ground distance, yes?) a multilook product?
Thank you for your support
HI,
Single look imagery, usually SLC data, uses the full spectral bandwidth of the received data. For multi-looked imagery, the bandwidth is split in to several looks. These looks are processed separately and incoherently summed. The fact that the bandwidth per look is reduced leads to a worse spatial resolution but the impact of speckle is reduced (but not removed). The incoherent summing means that the phase of the received signal is not preserved and so a multi-looked image is detected and so just contains amplitude values (and not complex as is the case for SLC data).
Ground range data has been resampled from the slant range direction using a suitable ellipsoid so that the range pixels are now along the ground. Usually the resampling is such that the pixel size is the same in ground range and azimuth (i.e. square pixels). This conversion to ground range is a separate processing step from the multi-looking but it is usually performed for multi-looked detected data.
Peter
2 Likes
Thank you for the clear explanation, it makes sense now.
This Sentinel tutorial would have you believe that multi-looking is the same as ground range projection, hence my confusion:
https://sentinel1.s3.amazonaws.com/docs/S1TBX%20SAR%20Basics%20Tutorial.pdf
Cheers.