There are three primary parts of the page:
- The Image
- The Work Area
- The Archive
Each serves an important purpose in the flow of composing alt text.
The primary purpose of the image area is displaying the image as you work on your description but
it also supports Optical Character Recognition, or OCR. OCR extracts readable text from images, but it
doesn't do a perfect job, and often sees random letters that aren't there in things like photographs.
In the main menu is a button that allows you to change the language being extracted.
The Work Area
The work area is below the image, and it's where you'll write descriptions. When you first open the
snippets will be there, you'll have to either use OCR on the image to extract text, or you can hit the
➕button on the left side to create a blank description. If it's too small, you can make it bigger by
dragging the bottom right corner of the text area.
You might have a bunch of snippets, especially if you're using OCR. Once you're happy with the
description(s) you've made, you can either click the 💾 on the snippet to save just one, or the one to
right will save all the current in-progress descriptions. If you don't want a snippet, you can hit 🚮 on
to delete it.
Any descriptions you save are moved into...
The archive has two main parts, the search bar and the results. When you first open the page, the
will be closed. You can enter search terms to find and display, but keep in mind the search operates on
When you upload an image, the archive is filtered down to descriptions written for that image. If you'd
to toggle that, you can hit the 🌠 button.
If you'd like to edit an item in your archive, click its 📝 button, and it will appear in the work area.
Make any changes and then hit 💾 to save them to your archive. If you want to discard the changes but
the old version, you can just hit 🚮 in the work area.
Set Max Length on a description to get a view of it split on whitespace. If the description is too long
be attached to a single image, you can click 🌠 next to each section to get an "Alt Text Continued"
That image will be localized if possible.