search

uw-cryo / asp_plot

Scripts and notebooks to visualize output from the NASA Ames Stereo Pipeline
BSD 3-Clause "New" or "Revised" License
4 stars 0 forks source link

vwip interest points on top of aligned (i.e. map-projected) L and R images (probably just use L_sub / R_sub) #8

Open bpurinton opened 5 months ago

bpurinton commented 4 months ago

Not sure this is needed. There is already functionality to plot the .match files, and the ASP docs note:

.vwip files (intermediate files for creating .match files).

So the .vwip interest points are the original raw interest points on each image that are filtered down to the single .match points file of the interest points found in both image.

There might not be a strong use case at the moment of implementing vwip plotting functionality.

If we do want it, then port the functionality from: https://github.com/friedrichknuth/bare/blob/3e610858be44418fcf334d760e853ce7868be16d/bare/plot/plot.py#L189

dshean commented 3 months ago

OK, fine for now with other priorities.

The motivation here is to see the interest points on each of the original images (to check distribution), and then to assess the subset of those interest points was successfully matched (and distribution). If we only have the match points, and there are issues with match point number and distribution, we don't know if the issue was due to poor matches, or the fact that there were no interest points to begin with in a particular area of the input images (e.g., due to limited texture, clouds, water).

Matches are more important, and when things work well, that's all we need. It's the cases that fail where we want to check the original ip.

bpurinton commented 3 months ago

Good points, I think it will be a nice feature to have, but as you said, other priorities first (i.e. going to start working on the ICESat plotting soon)