Opened 14 years ago
Closed 14 years ago
#347 closed bug (fixed)
DEM generation script fails on long lines
| Reported by: | benj | Owned by: | iopa |
|---|---|---|---|
| Priority: | alpha 4 high | Milestone: | |
| Component: | PML utilities | Keywords: | |
| Cc: | benj | Other processors: |
Description
The script used to generate DEMs for UK flightlines (nextmapdem.sh) currently only takes into account the start and end position of the aircraft on each line (because it reads these from the header files). This means that on long lines where there is an entire BNG tile between the start and the end it may not include DEM data for the whole flightline.
Possible solutions:
- (Preferred): Read SBET file in (using sbet_handler.py, unless you've got a pressing reason to do it another way), track position of aircraft and do maths on swath width to work out which BNG tile the aircraft is over (ensuring it's in the UK as well).
- (Possibly easier): Adapt existing nextmapdem.sh script to be more intelligent about what tiles might be in the middle. May be easier but might be complicated by corner case and/or case where rolls over tile edge (9->0).
Change History (2)
comment:1 Changed 14 years ago by iopa
comment:2 Changed 14 years ago by iopa
- Resolution set to fixed
- Status changed from new to closed
Note: See
TracTickets for help on using
tickets.
Choosing solution 1, I have created a script to replace, in essence, gettiles.sh, which is the source that nextmapdem.sh used to be told which tiles to load. I have tested my script demgen.py with a copy of nextmapdem.sh, by replacing gettiles.sh with the new script and a whole, continuous DEM file is successfully generated in the same manner as before (covering the the whole of the london flight lines, which were used as test samples).
What remains before it is released is for it to be documented and commented (more of an early morning job).