Build a searchable index of images by looking at the text inside the images.

Perfect for your unsorted meme directory where you just want to recall that one image that talked about a "bulletsponge" or something like that.

Built during a lab day at work. Not to be taken too seriously.

You index a directory of images and the text that can be recognized in the image is saved in a metadata file under .miui.

You can then search inside these metadata files and map the metadata filename back to the original filename. The built-in search uses fzf.

For indexing:

• tesseract - OCR software. Must also have the English dataset installed.
• imagemagic - To preprocess the images during indexing.

For built-in search tool:

• fzf - The fuzzy-finder used to present the search interface.
• kitty - The only supported terminal right now in order to get image previews. (Easy to fix!)

Licensed under Apache License 2.0.

Copyright © Magnus Bergmark 2019