1. Home
  2. Glossary
  3. Computer vision
AI glossary · Media & modalities

Computer vision

Computer vision: Computer vision is the field of AI that lets software interpret images and video: recognizing objects, reading text, detecting defects, tracking movement, or describing a scene. It powers everything from phone photo search to warehouse inspection cameras.

A computer sees an image as a grid of numbers. Computer vision models, usually deep neural networks trained on millions of labeled images, learn to turn those numbers into meaning: this is a forklift, that is a barcode, this weld has a crack, this receipt total reads $84.12.

Common workplace forms include optical character recognition (OCR) for scanned documents, object detection for inventory and safety monitoring, quality inspection on production lines, and image classification in insurance claims. Today's multimodal assistants fold general-purpose vision into chat, so you can upload a photo and ask questions about it.

Vision systems fail in predictable ways: unusual lighting, angles, or items they were not trained on. They can also raise privacy questions when pointed at people. If a system will make decisions about employees or customers from camera data, involve legal and HR before deployment.

Example at work

A regional distributor installs cameras at the receiving dock. A computer vision model reads carrier labels, counts pallets, and flags visibly damaged cartons as they come off the truck, creating a timestamped photo record. Disputes with carriers now start with evidence instead of memory.

Why it matters

Any process that currently depends on someone looking at something, whether counting, checking, reading, or matching, is a candidate for computer vision. Knowing the term helps you spot those opportunities and ask vendors the right questions about accuracy and failure modes.

Related terms