Show your support for community science. Donate to Public Lab »

Public Lab Research note


Comparison of 4 versions of imagery, moire effects and resampling

by warren | June 02, 2011 02:43 02 Jun 02:43 | #315 | #315

Check out this image of Wilkinson Bay, Louisiana, by the USGS. Clockwise from top left:

  • Geotiff from USGS opened in Photoshop
  • JPG from USGS opened in Photoshop
  • Geotiff from USGS opened in Preview -- notice it's apparently reading one of the bands as the alpha channel; cool!
  • TMS tiles served through TileStream, generated with raster2mb from the USGS geotiff

I suspected something funny with the upper right image; notice there's some checkerboard patterning due to antialiasing (a moire pattern) in the lower left of that image. I wanted to go back to the source to confirm that it wasn't just from the JPG compression, or from a resize - I'd read that this image was originally 70cm imagery resized to 1m. The original Geotiff has much better color, and while I do see some moire, it may actually be from the waves on the ocean surface interfering with the pixel resolution! I dunno, but I'm glad I went back to the source.


0 Comments

Login to comment.