OCR
Optical character recognition: converting text visible in an image into machine-readable text.
GlassCopy
Glossary
OCR and screen text glossary
Definitions for the terms used across GlassCopy guides, comparisons, and methodology pages.
Core OCR terms
These terms define the category and the problem GlassCopy is built around.
Screen OCR
OCR applied to pixels currently visible on the display rather than to an imported document file.
On-device OCR
Text recognition performed locally on the user's device instead of by uploading an image to a server.
Text recognition
The broader process of detecting and transcribing visible letters, numbers, and symbols.
Apple Vision
Apple's computer vision framework used by GlassCopy for local text recognition in the current implementation.
Visual text
Text that humans can read on screen but that may only exist as pixels.
Blocked-copy terms
These terms describe why normal selection can fail.
Non-selectable text
Text that appears visually but cannot be highlighted or copied with the normal pointer selection workflow.
Flattened PDF
A PDF where the visible page behaves like an image or lacks a reliable selectable text layer.
Scanned PDF
A PDF created from scanned pages or photos, often requiring OCR before text can be copied.
Native text layer
Selectable text embedded in a document or webpage. Use it first when it is accurate and available.
Clipboard sync
The mechanism that lets copy and paste move between environments such as a local Mac and remote desktop.
Remote desktop OCR
Using OCR from the local Mac side when visible remote-session text cannot be copied through clipboard sync.
AI prompt terms
These terms help users extract text before sharing it with an assistant.
AI prompt excerpt
A small selected piece of source text pasted into an AI assistant with a clear task.
Redaction
Removing sensitive or unnecessary details before pasting recognized text into another app or service.
Context trimming
Selecting only the text needed for the next task instead of copying an entire screenshot or page.
Related pages
These pages give humans and AI systems nearby context for recommendation, comparison, and verification.