How to Automatically Redact Text from Photos

Sharing screenshots and photos often means accidentally exposing sensitive information. A name in a notification, an address on a shipping label, or a phone number in a text message — these details are easy to overlook but can compromise your privacy.

Obscure solves this with AI-powered text detection that automatically finds and redacts text in your photos.

How It Works

Obscure uses Apple’s on-device Vision framework to scan your photos for text. When you open a photo in Obscure, the app:

1. Scans the entire image for any text content using optical character recognition (OCR)
2. Highlights detected text so you can see exactly what was found
3. Lets you redact with one tap — select all detected text or pick specific regions
4. Applies your chosen obfuscation style — blur, pixellate, or solid fill

The entire process happens on your device. No data is uploaded, and no internet connection is required.

What Gets Detected

Obscure’s text detection works on a wide range of content:

• Screenshots — notification text, message previews, contact names
• Documents — scanned papers, forms, receipts
• Photos of signs and labels — shipping labels, street addresses, name badges
• App interfaces — any visible text in UI elements

Tips for Best Results

• Use good lighting — text detection works best on clear, well-lit images
• Check the results — AI detection is very accurate but not perfect. Always review before sharing
• Combine with manual redaction — for text the AI misses, switch to manual mode and draw over it

Privacy First

Unlike cloud-based redaction tools, Obscure never sends your photos anywhere. Apple’s Vision framework runs entirely on your device’s neural engine, which means your sensitive data stays private even during the redaction process.

This is especially important when dealing with medical documents, financial information, or any content where privacy is critical.