2

I'm looking for an easy to use panorama stitcher for Linux that doesn't have too many dependencies.

I've tried Hugin, but it failed me: often didn't even match the images – and even if I added a bunch of control points manually, generating a useful result image was a mess and much too complicated. Photos taken with a smartphone often lack the lens information Hugin wants to know before loading, and I've got no idea where to get those details from (guess that's the main issue here for me, which then cuases the "bad results").

The one built-in to Gimp (formerly distributed separately as "Pandora") can only be used for real simple things, and otherwise fails e.g. with the perspective.

Took a look at fotoxx – but that wanted to first index all my images, which I do not want (what for, if I only want to create one panorama? I don't want the stitcher to manage my entire collection).

I'm currently using Linux Mint Cinnamon, so please no KDE apps (I don't want to draw in half of the KDE desktop just for this task). Being in the standard repositories (or even a PPA) would be a big plus. Should be free (as in "free speach" and "free beer").

Note: When I say "panorama" I don't mean 360° but rather 2-5 images to be joined ("wide shot").

Improve this question

2 Answers 2

Reset to default
1

Not an exact match, but a substitute that does the job for photos I've created with my Android smartphone anyway:

Bimostitch Panorama Stitcher matches all my requirements (except for the OS).

  • Easy to use: Select the photos from the gallery, tap OK, done. A progress bar in the notification area lets you see how far it already got. When done, the panorama is stored on the internal SD card in the Pictures/Panoramas/ folder, and shown on the app's initial page.
  • Not too many dependencies: a single app with ~6 MB .apk size. Requires Android 4.0 or higher – which should be easy to match.
  • Linux: No. This is an Android app.
  • Free: as in "free beer" – but not as in "free speach".

Bimostitch Bimostitch
Bimostitch (source: Playstore; click images for larger variants)

What I haven't yet tried is what the second screenshot shows: instead of running Bomostitch on "full automatics", you also can have it show you a preview where you can adjust things. And of course you can play with the settings – to e.g. chose the maximum image resolution, whether original images should be deleted after processing, and more.

I've not used it much yet (as indicated, I only need such a software occasionally) – but what I needed it for now it processed just fine.

Improve this answer
1

I have always had excellent results from Hugin so suspect what you need to do is to address the lens information from your smartphone(s) and add that to the images.

Finding the lens information for a given smartphone, assuming that it is one that fails to include the information, is for most smartphones simply a matter of going to the manufacturers web site &/or your user manual.

Once you have the information for your phone(s) you need to embed that information into each picture, this lives in the EXIF information, and the tool for that is exiftool which is a free, gratis & open source, cross-platform, tool for extracting and updating the EXIF information in a wide range of media files. You can easily create a bash script to add the relevant EXIF information, (Focal Length & FoV seem the most likely).

The other issue is training yourself in how to take photographs that are suitable for stitching together.

  1. Stay on one spot, just turn - this avoids introducing parallax errors.
  2. Take the shots over a quickly over a short time span - this minimises the lighting changes.*
  3. Ensure plenty of overlap between pictures - this allows easier stitching.
  4. Try to have elements with strong vertical & horizontal components in each overlap area - makes alignment easy
  5. Try not to vary any settings while taking the shots, especially zoom.
  6. Never use digital zoom.
  7. If the lighting varies drastically across the final picture try to get the lighting transition near the middle of at least one shot - sometimes it works better to make a lightside and darkside panorama separately then stitch those together this is easier if your transition is near a mid-frame.
Improve this answer
2
  • It might well be all due to that missing detail, and thanks for your answer – but what you suggest doesn't meet the "easy to use" part. I know how to shot photos suitable for panoramas incl. overlapping and exposure (thanks for the guide, though). Nevertheless: If that lens info were that crucial, how comes other tools don't need it? KDE has a panorama maker which I've used in the past, and it didn't ask for that (using the same photos Hugin complained about). // So thanks again, but that approach is too complicated (too many steps involved), and the websites not always have the lens details.
    – Izzy ♦
    Jul 30, 2017 at 13:20
  • 1
    Maybe this tutorial for stitching flat scanned images together might be useful: hugin.sourceforge.net/tutorials/scans/en.shtml
    – MrWhite
    Sep 23, 2017 at 12:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.