Pruning my collection - Gwenview and Geeqie

Processing
1

I have just put together a replacement for my old QNAP, using a Raspberry Pi and Open Media Vault.

When I came to populate it, I realised how much crap I had lying about. Collections that I had always meant to prune, but never quite got around to it.

I used fdupes to search for duplicates across multiple directories with the aim of then looking at the results in an image viewer. I am a KDE user, so I started looking at my collections in Gwenview, but it really is appalling slow at loading raw files. Eventually I used Geeqie.

Several things:

  1. What are others using to find duplicate images?
  2. What are others using to view duplicates and prune them?
  3. Geeqie is lightning fast loading lots of raw files, which leads me to think it is using the GPU on my system, while Gwenview is not. I have no idea whether Gwenview can use the GPU, and there doesnā€™t seem to be any otpion to make it do so. I will raise this on the KDE forum, but any input from users here would be welcome.
1 Like
2

For finding bit-identical dupes, I use dupeGuru. It also has a picture mode, but I havenā€™t tried it.

digiKam also has a duplicate finder feature, based on visual similarity. Again, havenā€™t really tried it.

I donā€™t know Gwenview, but itā€™s possible that it renders the raw data (using dcraw or libraw), while Geeqie probably shows the embedded JPEG preview.

2 Likes
3

It uses the embedded JPG. Donā€™t know about Gwenview.

4

I use it for finding dupes, even different file sizes or formats. It is not completely perfect, and will occasionally flag very similar pics as dupes. There are parameters and they do change its behaviour.

I havenā€™t done this for ages, and since I started working primarily with raw, I know Iā€™ve got several different jpeg versions for many files. The raws themselves, as I store them, will always be unique, so I do not need to ā€˜viewā€™ them. I use qimgv as a fast and simple non-raw image viewer.

1 Like
5

Open up the terminal and type:-
sudo find / -name hello*.jpg
OR
sudo find / -name *hello.jpg
OR
sudo find / -name *hel.jpg

6

In Geeqie, File > Find Dulpicates

In digiKam, Tools > Find Duplicates

There is an app call dupeGuru

Another app called Czkawka

From the CLI there is fslint, rdfind, fdupes, rmlint

1 Like
7

Itā€™s probably enough to do this with user rights, rather than as root.
-iname (instead of -name) would be case-insensitive. But what exactly were you suggesting? I donā€™t think such a named-based search helps much with deduplication / image pruning.

1 Like
8

Usually I use the captioned command to search the duplicated files in my KDE desktop PC. Dolphin browser has the "search featureā€™. Searching for the files ā€œsort by sizeā€ would be the another way in Dolphin.

9

But for that you need to know (part of) the name of the file, and youā€™d have to run the command for each pattern (partial filename).

10

In Dolphin browser, search for file suffix i.e. jpg, CR2, NEF etc and then using ā€œsort by sizeā€ to look up the duplicated file.

11

And one would probably not want to search oneā€™s entire /.

Unix/Linux tools are wonderful. I have not even scratched the surface of the find command, and that includes 15 years as a systems manager. But tools for the job are better. dupeguruā€¦ fslintā€¦

(Didnā€™t fslint fall by the wayside and get reborn with a new name?)

1 Like