Le vendredi 23 octobre 2015 11:26:11, Helmut Kudrnovsky a écrit : > hi, > > tested with > > OSGeo4W: gdalinfo --version > GDAL 1.11.3, released 2015/09/16 > > I have several raster ecw files; gdalinfo shows for all the files the same > extent/corner coordinates, although the ecw files are some kind of mosaic > over a larger area. (Also QGIS and GRASS GIS show them in one place > (overlayed one to another)). > > example: > > the content of the world file: 6617-04.eww > > 0.2 > 0.0 > 0.0 > -0.2 > 653750.1 > 169999.9 > > gdalinfo says: > > Driver: ECW/ERDAS Compressed Wavelets (SDK 5.0) > Files: 6617-04.ecw > 6617-04.eww > 6617-04.ecw.aux.xml > Size is 6250, 5000 > Coordinate System is: > PROJCS["MGI_BMNM34", > GEOGCS["GCS_MGI", > DATUM["Militar_Geographische_Institute", > SPHEROID["Bessel_1841",6377397.155,299.1528128]], > PRIMEM["Greenwich",0.0], > UNIT["Degree",0.0174532925199433]], > PROJECTION["Transverse_Mercator"], > PARAMETER["False_Easting",750000.0], > PARAMETER["False_Northing",-5000000.0], > PARAMETER["Central_Meridian",16.333333333333], > PARAMETER["Scale_Factor",1.0], > PARAMETER["Latitude_Of_Origin",0.0], > UNIT["Meter",1.0]] > Origin = (0.000000000000000,69.444444444444443) > Pixel Size = (0.013888888888889,-0.013888888888889) > Metadata: > COLORSPACE=RGB > COMPRESSION_RATE_TARGET=10 > VERSION=2 > Corner Coordinates: > Upper Left ( 0.0000000, 69.4444444) ( 6d51'43.49"E, 44d44'46.05"N) > Lower Left ( 0.0000000, 0.0000000) ( 6d51'43.85"E, 44d44'43.83"N) > Upper Right ( 86.8055556, 69.4444444) ( 6d51'47.38"E, 44d44'46.37"N) > Lower Right ( 86.8055556, 0.0000000) ( 6d51'47.75"E, 44d44'44.15"N) > Center ( 43.4027778, 34.7222222) ( 6d51'45.62"E, 44d44'45.10"N) > Band 1 Block=256x256 Type=Byte, ColorInterp=Red > Description = Red > Overviews: 3125x2500, 1562x1250, 781x625, 390x312, 195x156 > Band 2 Block=256x256 Type=Byte, ColorInterp=Green > Description = Green > Overviews: 3125x2500, 1562x1250, 781x625, 390x312, 195x156 > Band 3 Block=256x256 Type=Byte, ColorInterp=Blue > Description = Blue > Overviews: 3125x2500, 1562x1250, 781x625, 390x312, 195x156
Helmut, Looking at the ECW driver code, if the .ecw file itself has an internal georeferencing, then the world file .eww is ignored, which looks consistent with the above report. This is a debatable behaviour. I think this has been discussed quite a few times. Maybe a config option to override the behaviour would be useful. A potential workaround is to edit 6617-04.ecw.aux.xml with the following content derived from the .eww. .aux.xml overrides the internal georeferencing of the .ecw. <PAMDataset> <GeoTransform>653750,0.2,0,440720,0,-0.2</GeoTransform> </PAMDataset> Even > > and the layer properties in Arcgis says: > > <TreeList> > <Raster_Information> > <Columns_and_Rows>6250, 5000</Columns_and_Rows> > <Number_of_Bands>3</Number_of_Bands> > <Cell_Size__X._Y_>0,2, 0,2</Cell_Size__X._Y_> > <Uncompressed_Size>89,41 MB</Uncompressed_Size> > <Format>ECW</Format> > <Source_Type>Generic</Source_Type> > <Pixel_Type>unsigned integer</Pixel_Type> > <Pixel_Depth>8 Bit</Pixel_Depth> > <NoData_Value></NoData_Value> > <Colormap>absent</Colormap> > <Pyramids>level: arbitrary, resampling: Nearest Neighbor</Pyramids> > <Compression>ECW</Compression> > <Mensuration_Capabilities>Basic</Mensuration_Capabilities> > <Status>Permanent</Status> > </Raster_Information> > <Extent> > <Top>170000</Top> > <Left>653750</Left> > <Right>655000</Right> > <Bottom>169000</Bottom> > </Extent> > <Spatial_Reference>MGI_BMNM34 > <Linear_Unit>Meter (1,000000)</Linear_Unit> > <Angular_Unit>Degree (0,0174532925199433)</Angular_Unit> > <False_Easting>750000</False_Easting> > <False_Northing>-5000000</False_Northing> > <Central_Meridian>16,333333333333</Central_Meridian> > <Scale_Factor>1</Scale_Factor> > <Latitude_Of_Origin>0</Latitude_Of_Origin> > <Datum>D_MGI</Datum>< > /Spatial_Reference> > </TreeList> > > any ideas? > > thanks > > > > ----- > best regards > Helmut > -- > View this message in context: > http://osgeo-org.1560.x6.nabble.com/ecw-file-wrong-extent-corner-coordinat > es-tp5232129.html Sent from the GDAL - Dev mailing list archive at > Nabble.com. > _______________________________________________ > gdal-dev mailing list > [email protected] > http://lists.osgeo.org/mailman/listinfo/gdal-dev -- Spatialys - Geospatial professional services http://www.spatialys.com _______________________________________________ gdal-dev mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/gdal-dev
