stripes in results

Added by Lucas Johnson about 1 month ago

I'm using GAMMA to do offset tracking over Pine Island Glacier in Antarctica. For some image pairs, where the pair only partially intersects the fast moving portion of the glacier, I get intermittent stripes (alternating high and low displacements/velocities) in the results. Pairs centered on the fast moving portion of the glacier never have these issues...

The stripes don't seem to be aligned with burst/swath boundaries. Could these be "edge artifacts"? Or something else? Any ideas would be appreciated.

Thanks!

clipboard-202606221718-7ymxo.png (93.2 KB) clipboard-202606221718-7ymxo.png

Replies (1)

RE: stripes in results - Added by Lucas Johnson about 1 month ago

Turns out I fixed this by more aggressive masking (ccp threshold 0.05 -> 0.15) and simpler gap filling (8-nearest neighbors rather than spring-analogy) between offset tracking rounds 1 and 2 (offset_pwr_tracking and offset_pwr_tracking2). It seems generating a smooth surface to seed the second round is important. Still open to any thoughts others might have on this.

Thanks!

(1-1/1)

Project

General

Profile

GAMMA Questions