Enhancement: Handling workflows for extracting multiple images and indicating previous croped areas

[brierjon]

Currently the workflow to extract images from images such as: https://commons.wikimedia.org/wiki/File:House_of_Representatives_for_the_State_of_Tennessee_1891-1893.jpg or pdf with multiple pages like: https://commons.wikimedia.org/wiki/File:Today%27s_pictorial_story_of_Alexander_Campbell_and_the_Christian_churches_(IA_todayspictorials00john).pdf are slowed significantly. The current workflow has a user start over with each crop and previous crops are not visible to see what has already been extracted. This can lead to human error either skipping or double cropping the same object.

Has there been consideration of allowing users to crop multiple objects then put through the naming process? Ideally this would include adding a categorization selection of which categories should remain with the crop vs the source. With multiple extracts like these could lead to many relevant categories per person added which isn't applicable to each cropped photo.

Has there been consideration to overlay which crops have already been done by highlighting the area cropped previously?