#3379 closed defect (fixed)
gdalinfo v1.6.3 is incorrectly reading the georeferencing on pix files in UTM zones
Reported by: | abarton | Owned by: | warmerdam |
---|---|---|---|
Priority: | normal | Milestone: | |
Component: | OGR_SRS | Version: | unspecified |
Severity: | normal | Keywords: | pcidsk |
Cc: | ysiddiqui@…, gaopeng, pvachon |
Description
It appears that gdalinfo v1.6.3 is incorrectly reading the georeferencing on pix files in UTM zones. Specifially the false northing is coming across as 10,000,000 instead of 0 and placing it in the southern hemisphere.
A sample can be found at ftp.i3.com using the fwarmerdam login. It is named georeferencing_sample.pix.zip.
Change History (4)
comment:1 by , 14 years ago
Cc: | added |
---|---|
Component: | default → OGR_SRS |
Keywords: | pcidsk added |
Milestone: | → 1.6.4 |
Status: | new → assigned |
comment:2 by , 14 years ago
After a bit more discussion with Phil, I have applied a small fix in trunk (r18735) that makes the MGRS zone interpretation a bit more resilient in the case where something not intended to be an MGRS code gets picked up as an MGRS code.
comment:3 by , 14 years ago
Resolution: | → fixed |
---|---|
Status: | assigned → closed |
Note:
See TracTickets
for help on using tickets.
This appears to be a problem introduced in r18096 by Phil in an attempt to provide interpretation of MGRS codes to indicate southern hemisphere. In this case the string is "UTM 13 D000" and ' ' is picked up as the MGRS code and misinterpreted as being southern hemisphere since it is not northern hemisphere.
This problem affects 1.6.3 and 1.7.0 both old and new PCIDSK drivers.
I believe I have corrected it in trunk (r18730), 1.7 (r18731), 1.6 (r18732), and 1.6-esri (r18733).
Phil could you review the change? I simplified the code quite a bit assuming that PCI strings from files are well formed, and that the MGRS zone code would always be in position 10 if set.