Duplicate Image Finder

Duplicate image finder uses image hashing to find similar/duplicate images in your local storage. All you gotta do is

1. install,
2. run (will setup the database with table) if no configuration is provided,
3. run specifying which directory to look for images, and finally
4. run asking it to show duplicate/similar images.

Please note that it is a prototype. Please use at your own discretion.

For example:

# 1. installing
python3.9 -m pip install --user duplicate-image-finder

# 2. show help
duplicate-image-finder --help

# 3. add directory images and calculate hashes using 4 threads
duplicate-image-finder --add <directory> --parallel 4

# 4. show the duplicate/similar images found in your browser
duplicate-image-finder --show

Running 4 will result in opening a browser that shows duplicate/similar images. If you click on delete, it will be moved to .Trash folder.

Requirements

Lots, but all of them can be installed as dependencies as long as you are using python3.9. Unfortunately, some of its dependencies have not been made available in python3.10 yet, so we are stuck there.

Poetry

Installing dependencies

poetry install

Running

poetry run python duplicate_image_finder/duplicate_finder.py --show

Testing

poetry run pytest

etc.

This duplicate image finder source code is inspired/partially copied from https://github.com/philipbl/duplicate-images.git.

Significant changes from the referred version are:

1. moved from mongodb to sqlite
2. Is probably better in terms of finding similar images (or perhaps I misunderstood the previous code)

Concepts/Technologies I learned/tried to learn while doing this:

1. poetry for dependency
2. pytest for unit test
3. pysqlite3 for database
4. concurrency for performance
5. imagehash for perpetual image hashing for finding similarity
6. grouping CLI arguments in python (mutually exclusive, etc) using argparser