Turning photos into searchable text

Whiteboards get erased. Documents get filed. SempreNotes’s photo path turns what you can see into something you can ask about later — with text extraction running on-device through Apple’s Vision framework.

Capturing

1. Tap the photo button, then choose the camera or your photo library.
2. Take or pick the photo.
3. SempreNotes runs OCR on-device and shows you the extracted text alongside the image.
4. Commit to store both as one memory with source photo.

Honest OCR

OCR is imperfect, and SempreNotes is honest about it rather than confident about it:

• Partial reads — if glare washes out half the whiteboard, SempreNotes stores what it recognized and flags low-confidence regions. It never invents the missing text.
• No text at all — a landscape photo stores with an empty-text flag and a notice that nothing was extracted. That is a normal state, not an error.

Searching what you captured

Extracted text is indexed locally and in iOS Spotlight. Ask the brain “what were the pricing tiers on that whiteboard?” and the answer cites the photo memory it came from.