Hi,
I am doing the backscattered value extraction for TerraSAR-X muti-temporal dataset including 16 of which for the descending orbit. The study area is flat because it consists of paddocks.
My workflow is:
1.Calibrating
2.Multilooking
3.Subset
4.Co-registration
5.Multi-temporal Speckle filtering
6.Terrain-correction
My problem is that after step 5, some of image acquisition dates in which black individual pixels are “created”, compared to the corresponding one in stack, they can be shown below:
This is one of subsets of the stack after it being multi-temporal filtered (Lee 5x5, tried all the methods, this one preserves good features)
As you can see, the black pixels are within light areas.
This is the one with same date as above in the stack without multi-temporal filtering.
Sorry I am a new user who can just only post two pics right now.
In terms of the things above, I’m wondering am I doing right? Is these 'black individual pixels’left a normal thing?
or does this problem will affect the intensity extraction for these paddocks?
Thanks for your reading and helping.
