1

I have thousands of photos in multiple directories. Any tool or idea to organize them properly will be helpful.

Here is the details:

I have been ignorant all along about organizing those until recently when I tried to search a particular picture. However, I am using a bash script to find and delete duplicates (md5 hash data) and imagemagick's identify to extract EXIF:DateTimeOriginal. My image file names are like yyyymmdd_serialNo.extension. For example,

20210101_1.JPG

Most of the pictures made on my DLSR and iPhone have those info except for the ones which are very old. Even fdupes can't detect the duplicates of those. I could not but rely on my visual capabilities (a pain you can imagine). Any thoughts to improve what I am doing? I appreciate

15
3

There is an element of XY problem in the question. What you probably want is not organizing photos efficiently, but finding photos efficiently. Framed that way, deduplication has approximately zero benefit (if you're out of disk space a bigger disk is a better answer).

Framed as improving the finding of photos, any first step is pretty much a step in the right direction (or at least not a regression) whereas deduplication has a high potential for permanently making a picture unfindable...human fallibility works like that.

With pictures produced by digital cameras, EXIF data comes more or less for free, and so that's a place to start. There are many options for cataloging pictures. I happen to use Darktable for that because I also use it for editing pictures. But any catalog program that stores EXIF data in a database table when pictures are imported is worth considering.

The problem with EXIF data is that it is (in ordinary practice) entirely technical. It does not contain information about the contents from the standpoint of what we humans care about.

Tags are the most straight forward way to add information that matters in terms of what we do with photographs. The great thing about tags is that two similar tags on the same picture don't make a difference -- a picture with "dog", "dogs", and "canines" is easier to find. Even better tags are not mutually exclusive. A relational model with a field "color" won't allow "green" and "brown" and "blue" a the same time. The tags model does.

If you start tagging new pictures today, you have improved finding pictures. There's no need to go back and boil the ocean by organizing every single image. At least right away. Instead, images can be organized bit by bit as needed. Keep in mind that you will never want to find most of your images.

Just importing the images into a "cataloging" application will go a long way because EXIF typically contains date information and usually you will have an approximate timeframe for when the picture was made.

As your tagging grows, you will also have some idea of how your tagging practice has improved over time. (For me simpler and flatter is better).

1
  • Really good for recognizing the oblique problem of finding what you have vs. processing incoming. – scottbb Jan 18 at 21:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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