Skip to content

Image Occlusion

Image occlusion lets you hide labelled parts of an image and recall them during review.

It is useful for diagrams, anatomy, maps, pathways, and anything else where location matters.

The normal workflow is:

  1. Open Add flashcard.
  2. Choose Image occlusion.
  3. Paste an image or drag an image file into the modal.
  4. Open the image occlusion editor.
  5. Draw or generate masks.
  6. Save the card.

LearnKit also supports a dedicated Add image occlusion flashcard to note command.

[!NOTE] The full image occlusion editor is desktop-only.

The current editor supports:

  • rectangle and ellipse masks
  • moving and resizing masks
  • undo and reset actions
  • pan and zoom controls
  • image crop and rotation tools
  • text annotations
  • OCR-powered auto-masking

Auto-Mask looks for text-like regions in the image and creates masks for them.

This is useful for labelled diagrams, but it is still a draft step. You should expect to clean up the result manually.

In practice:

  • clear, high-contrast labels work best
  • existing masks are kept
  • generated masks can be moved, resized, regrouped, or deleted before saving

Image occlusion supports two save modes:

ModeWhat it does
SoloCreates one review target at a time so each mask or group is tested separately
AllHides all masks together on the same card

Each mask can have a group key.

Masks with the same group key behave as one unit. This is useful when several labels belong together and should be hidden or revealed together.

During review, LearnKit shows the image with occlusion overlays.

After you reveal the answer, the app can either:

  • reveal the target group only
  • reveal all groups at once

Configure that in Settings -> Flashcards -> Image occlusion -> Reveal mode.

Image occlusion files are stored separately from the review setting above.

You can configure these in Settings -> Data & Maintenance -> Attachment storage:

  • Image occlusion folder
  • Delete orphaned image occlusion images

The default folder is Attachments/Image Occlusion/.

  • anatomy diagrams
  • labelled maps
  • flowcharts with named parts
  • processes where a visual landmark matters

If the content is mostly text and order matters more than location, Ordered Questions or Multiple Choice Questions are usually a better fit.

Last modified: 30/03/2026