Has anyone tried mosaic jp2000 imagery in ERDAS, ArcGIS , ENVI or any softwares that they have accomplished ?

I have three jp2000 files that is total about 28.1 GB that I want to mosaic with ?

  • Should be no trouble for ERDAS so if you have is just try. Can't say anything about ArcGIS and ENVI. GDAL is good too if it has a good JPEG2000 driver. – user30184 Dec 2 '14 at 21:58
  • Ok. I can give that a try. – PROBERT Dec 2 '14 at 22:01
  • 1
    The trouble with mosaicing JP2K is the not-quite-nodata created by a lossy compression will overwrite adjacent images (unless the compression is lossless). I would definitely go with ERDAS as you can define your 'data area' for a smooth seamline and it is considerably faster than Esri at raster processing operations. GDAL is fast too but lacks the tools by default for a masked mosaic operation - if you're considering GDAL then perhaps the QGIS interface (and some plugins) would be the way to go. Yes, I have done this in GDAL but images were all data to the edge. – Michael Stimson Dec 2 '14 at 22:22
  • @MichaelMiles-Stimson - Good to hear..Yeah what I am the most concern is that once you mosaic these jp2 files it has to uncompressed first and then mosaic and then compressed again and save as a jp2000 again ? Do I make myself clearly ? – PROBERT Dec 2 '14 at 22:26
  • Yes, that's the way I did it. I used GDALbuildvrt to make a single image but compressing direct from VRT with JP2K tiles to single JP2K was very slow and crashed. The way I did it was VRT with JP2K tiles -> Uncompressed GeoTiff -> JP2K. The uncompressed image was over 5 TB (I had to do it on Adaptec RAID as no single disc was big enough).. you could try LZW/Deflate compressed GeoTiff but it's still going to be very big. If you have access to ERDAS try that first, it has its own drivers which may not have the same problems going compressed to compressed image. – Michael Stimson Dec 2 '14 at 22:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.