IC image stitching as a basic level is similar to stitching together a large number of scanned images. Depending on how you collect the images though you also might have to (at least want to) correct lens defects, roll/pitch, and other parameters. Additionally, due to the very large sizes involved you may find that software breaks down as you start approaching the gigapixel (GP) image size.

This page primarily focuses on Hugin / libpano




I (JM) use this to stitch my images. Its intended to help automate overall workflow and accelerate large stitch jobs.

Gathering control points

The best open source way to do these is probably autopano-sift-c (or the surf version if you are commercial). However, I've found that Andrew Jenny's free but closed source version tends to do a lot better. I suspect that its related to point validation but don't really know. pr0nstitch takes cares of the hoops you'd need to go through to get the latest version of it working well through WINE.

Google maps view

Some options:

  • pr0nmap, part of pr0ntools
  • whitequark tool (FIXME: link)

See some examples here http://siliconpr0n.org/map/

People tend no to like to download huge image files and some people won't have the computer hardware to open them. Google maps view breaks the huge images into many smaller tiles so that they can be accessible to a wider audience (and save server bandwidth too!).

pr0ntools provides tools to convert an existing image to Google maps (pr0nmap) as well as directly stitching a panotools .pto project to tiles. In practice this tends to be faster than stitching to a full sized image and then breaking out because it avoids creating large intermediate images that tend to eat cache / RAM for breakfast.

software/stitching.txt · Last modified: 2016/11/13 01:47 by mcmaster
Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution 4.0 International
Recent changes RSS feed Donate Powered by PHP Valid XHTML 1.0 Valid CSS Driven by DokuWiki