28
\$\begingroup\$

I'm looking for a desktop search engine for finding images on my local hard drive with high visual similarity. I've only found paid for ones so far. It could be Linux or Windows software; either would be helpful. A Google Desktop plugin would be nice!

My first use case is being able to organise all the processed versions of a photo based on it's original, or finding the original of a processed image.

Improve this question
\$\endgroup\$
4

7 Answers 7

Reset to default
11
\$\begingroup\$

ImgSeek is an open-source project that claims to do this.

DigiKam is apparently trying to add similar functionality.

Pixcavator is not open source, but there is a 30 day free trial.

If you're the DIY coder type, I wound up writing the DB-end of a system that allows phash similarity queries over a fairly large image corpus (10M+ images) in <10 ms/image.

It's part of a larger image deduplication system, and rather tightly coupled, but it's all online here.

Later Edit:

I wound up writing a custom PostgreSQL indexing extension to push fuzzy-image-searching into the database. It's open source here. It scales up to at least 30 million images without significant issue (I have ~32 million images in my index).

Improve this answer
\$\endgroup\$
5
  • \$\begingroup\$ I'm trying imgSeek but it keeps crashing when I try to add my entire collection , about 20k images. It might get there eventually though and looks quite good \$\endgroup\$
    – barrymac
    Commented Jul 18, 2011 at 15:52
  • \$\begingroup\$ Yeah. It was last updated in 2006, though. \$\endgroup\$
    – Fake Name
    Commented Jul 18, 2011 at 22:04
  • \$\begingroup\$ imgSeek has since been updated in April 2013. \$\endgroup\$
    – the_new_mr
    Commented May 12, 2020 at 8:21
  • \$\begingroup\$ ImgSeek just crashed during my first scan. Seems a bit unintuitive to use as well. Just my thoughts. \$\endgroup\$
    – TripleAntigen
    Commented Feb 21, 2021 at 12:57
  • \$\begingroup\$ Can the open source tools mentioned find a picture that has a part of it in other pictures? \$\endgroup\$
    – baptx
    Commented Apr 4 at 22:46
11
\$\begingroup\$

After looking for a while I found the perfect solution. This program does exactly what I wanted and it's free: http://www.duplicate-finder.com/photo.html

Improve this answer
\$\endgroup\$
1
  • \$\begingroup\$ This "Awesome Duplicate Photo Finder" worked better for me because it was simple enough. VisiPics seems powerful, but I couldn't quickly figure out how to use it. \$\endgroup\$
    – Ryan
    Commented Feb 15, 2018 at 23:19
7
\$\begingroup\$

I've been using visipics which i found out about on lifehacker which works very well. Visipics

This is the excerpt from the site

I recommend VisiPics (pictured above). It scans the actual photo content of each image file, and so can take some time to tally up its findings if you've got a huge, huge database, but you can fairly easily let it run in the background and do your normal browsing and work. Matches are shown together, a sliding scale of matching looseness is offered, and while it's a shareware app with some locked features, it offers enough as a freeware app to do some serious cleaning.

How do you get started with VisiPics? Choose your photo folder from the list in the top-middle row, hit the right-facing arrow with the "+" sign to add it to VisiPics' queue, then hit the Play-style button in the middle to start scanning for duplicates.

Note: If you've previously used Picasa, iPhoto, or other photo management suites to organize and edit your photos, you may find lots and lots of photo duplicates. That's because those applications create folders of "Originals" to allow you to undo changes you make to photos, along with caches of thumbnail images for fast browsing. If you're not using these apps anymore, and you're sure you won't want to undo those changes, go ahead and remove those folders and duplicates. Otherwise, keep them, and remove their folders from your searches.

Speaking of Picasa, Google's photo manager has a "Show Duplicate Photos" option in its Experimental menu, but that's exactly what it is—very experimental, and not all that clear about what you're deleting. You're better off going with an app that makes its findings clear.

Then there's also dupeguru from this answer on superuser

Improve this answer
\$\endgroup\$
1
\$\begingroup\$

Visipics is an image duplicate utility with an adjustable similarity comparison scale. You can look for very similar pictures or fairly dissimilar pictures and then delete whichever ones you want.

Improve this answer
\$\endgroup\$
1
\$\begingroup\$

Similar.pictures is a browser-based finder of similar photos on your hard drive I developed. It does not allow to delete files, only shows groups of similar images, also as a list of files in text format.

Improve this answer
\$\endgroup\$
0
0
\$\begingroup\$

Geeqie, an image browser for Linux and friends, has this functionality, but I've never made serious use of it, so can't vouch for how well it works. It's a great, extremely configurable image browser in general, though.

Improve this answer
\$\endgroup\$
-3
\$\begingroup\$

Try TinEye. They have in-browser plugins -- not desktop. See how it works for you. Also, Google Image Search is getting pretty good at this. It's worth taking out for a spin every so often.

Improve this answer
\$\endgroup\$
3
  • 4
    \$\begingroup\$ -1 : The OP specifically mentions he needs a local option. \$\endgroup\$
    – Fake Name
    Commented Jul 18, 2011 at 5:07
  • \$\begingroup\$ +-0: That information was added to the question after this answer was posted. \$\endgroup\$
    – blubb
    Commented Jul 18, 2011 at 8:15
  • 4
    \$\begingroup\$ @Simon Stelling - I considered that. The local information was added to the title after the question was asked. It was always in the question body. OP: I'm looking for a desktop search engine for finding images on my local hard drive with high visual similarity. \$\endgroup\$
    – Fake Name
    Commented Jul 18, 2011 at 22:05

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