Opened 9 years ago

Closed 9 years ago

#3338 closed defect (fixed)

Missing georeferencing

Reported by: gaopeng Owned by: warmerdam
Priority: normal Milestone:
Component: GDAL_Raster Version: 1.6.2
Severity: normal Keywords: HFA
Cc:

Description

The extent values should be in world space (top: 2898846.6, left: 598543.8, right: 598644.6, bottom: 2898758.4).

Attachments (1)

QB_b1_subset.img (210.7 KB) - added by gaopeng 9 years ago.

Download all attachments as: .zip

Change History (6)

Changed 9 years ago by gaopeng

Attachment: QB_b1_subset.img added

comment:1 Changed 9 years ago by warmerdam

Status: newassigned

Gao,

It appears the problem is that the georeferencing node is named "MapInfo?" instead of the normal "Map_Info". I can change the HFAGetMapInfo() code to search for any node of type Eprj_MapInfo if Map_Info does not exist. Does that seem reasonable to you, or does experience suggest anything else?

comment:2 Changed 9 years ago by gaopeng

Yes.

comment:3 Changed 9 years ago by warmerdam

Keywords: HFA added
Milestone: 1.6.4
Resolution: fixed
Status: assignedclosed

I have added this logic in trunk (r18605), 1.6-esri (r18606) and 1.7 (r18607).

I deem this a rare form of file and am not taking the change back into 1.6 or 1.5.

comment:4 Changed 9 years ago by gaopeng

Resolution: fixed
Status: closedreopened

The cellsize is still missling. It should be 0.6 x 0.6.

comment:5 Changed 9 years ago by warmerdam

Resolution: fixed
Status: reopenedclosed

It seems the pixelSize variable in this file uses subfields x and y instead of width and height as is normally the case. I have added a fallback to accept this in trunk (r18621), 1.7 (r18622) and 1.6-esri (r18623). However, I suspect that the file was generated via a non-standard process and I have little reason to believe there are many others like it.

With the above changes this file seems to work ok.

Note: See TracTickets for help on using tickets.