Coherence ccd

i am working on new topic in coherence change detection but i can’t resolve this problem in loss coherence

in the same descending orbit
best regard

please describe your problem a bit more.

as a user
to estimate the coherence change detection in this area we took the bridge as landmark witch is we don’t have the same pixels values for all the coherence images that create a wrong in CCD

Are all images acquired from the same track? Coherence is very sensitive for looking direction.

Also, have you scaled all four images to the same color range in the Color Manipulation tab? You should set 0 as minimum and 1 as maximum for all images you want to compare first.

I’d suggest also to check the scene names in ASF. Besides the track they use frame number, which I found to be different for some areas. So use only scenes with same frame number.

yes all the images in the same track “Relative Orbit 66”

now we scaled all the images min 0 mid 0.5 max 1

we used the same relative orbit 66 for this area

Thank you for clarification, this helps us understand your problem.

@mengdahl do you have experience with solid metal objects from your coherence-based projects which could explain the observed differences?

That is interesting, in one of the pairs the coherence is very high and in others lower and with changing spots of high and low coherence. I wonder if this can related to traffic patterns on different days of the week? If there are parked cars on the birdge and their position changes, coherence must go down.

What is the size of the bridge & coherence estimation window? A high resolution optical image of the bridge would be useful.

we are not interested on bridge in that area we took “the bridge” as landmark to estimate the coherence change detection “ccd” for all the area we must first equalify all images in the same pixels values

Maybe such a metal structure is not the best reference target to normalize the information between the images.

what should we do sir for that please?

Perhaps you could try some kind of image normalization to make them more comparable. Note that repeat-pass coherence is affected by environmental conditions so you cannot expect that the coherence of a certain area is stable if there has been no disturbance to that area. But if something is disturbed, the coherence should drop to very low values, close to those over water-surfaces.