Opened 15 years ago
Closed 8 years ago
#2651 closed enhancement (wontfix)
Transform Statistics
| Reported by: | rjc1 | Owned by: | warmerdam |
|---|---|---|---|
| Priority: | normal | Milestone: | |
| Component: | default | Version: | 1.5.3 |
| Severity: | normal | Keywords: | transform statistics |
| Cc: |
Description
It would be of great benefit to have an optional CreateTransform output structure (for all versions such as Affine or GCP) that provides the resulting "actual" values of (xscale, yscale, rotation, xoffset, yoffset, xcentroid of points, ycentroid of points, xflip = true (if image is flipped), yflip = true (if image is flipped)]. For the Affine transform and GCP transform (all orders), these values can be computed from the resulting polynomial terms. I am working on a NIST project with this requirement.
Change History (2)
comment:1 by , 15 years ago
| Milestone: | 1.5.4 |
|---|---|
| Priority: | high → normal |
comment:2 by , 8 years ago
| Resolution: | → wontfix |
|---|---|
| Status: | new → closed |
No patch sent so far, closing.
Note:
See TracTickets
for help on using tickets.
I'm surprised that these can be computed from higher order polynomials in a dependable way, and the requirement seems fairly specialized. If you produce a solution that you would like to offer as open source, please attach to the ticket, though I'm not sure if it would be integrated since it would be specialized to just one transformer.