Computer Vision is the field of AI focused on enabling computers to interpret and understand visual information from the world — images and video — the way humans use sight.
What it means in plain English
Computer vision teaches machines to make sense of pixels. Given an image, a computer vision system can identify what objects are present, where they are, what a scene depicts, or whether something has changed. Modern computer vision is powered by deep learning, particularly neural networks trained on huge collections of labelled images, which learn visual features layer by layer.
It covers a wide range of tasks, from simple image classification to detecting and tracking multiple objects in live video.
A simple example
When your phone unlocks by recognising your face, or a car’s safety system spots a pedestrian, that is computer vision — a model interpreting a camera feed and identifying what it sees.
Why it matters
Computer vision underpins face recognition, medical imaging, self-driving cars, quality inspection in factories, and much more. As cameras become ubiquitous, it is one of the most widely applied areas of AI.
Related terms
- Deep Learning — the technique powering modern computer vision.
- Object Detection — a core computer vision task.
- Convolutional Neural Network — a network design built for images.
Frequently asked questions
What can computer vision do?
It enables machines to interpret images and video — recognising objects and faces, reading text, detecting defects, guiding self-driving cars, and analysing medical scans, among many uses.
What AI powers modern computer vision?
Much of it is powered by deep learning, particularly convolutional neural networks and, increasingly, transformer-based vision models.
