Anki image occlusion examples: what works and what fails
Inspect the source figure, mask result, and review decision for three real patterns. Use the checks before spending time turning a diagram into Anki cards.
The 30-second preflight
- 1Can you select the label text in the PDF?
- 2Are there at least three useful labels?
- 3Does each label point to one clear target?
- 4Will the whole answer fit under one mask?
- 5Can you split a crowded plate by region?
Strong automatic candidateClean heart labels
A worked six-label anatomy example showing the spacing, text layer, and leader-line pattern that automatic Anki image occlusion handles well.
6 masks · source and authorization included
Inspect this example →
Automatic draft; manual review requiredDense vessel labels
A dense heart plate showing why automatic masks need a human review pass when leader lines cross and the label count exceeds a useful card.
3 masks · source and authorization included
Inspect this example →
Pass after visual reviewNo-leakage coverage
A before-and-after review standard for checking that every image-occlusion mask covers the complete label without hiding the diagram target.
6 masks · source and authorization included
Inspect this example →
Check your own labeled diagram
Run the anonymous demo first, then upload the full PDF only when the source has a usable text layer.