LEARN / TECHNOLOGY
Computer Vision
Computer vision uses computational methods to extract information from images and video.
How it works
Models process pixel data to perform tasks such as detection, segmentation, tracking, depth estimation or scene understanding.
Why it matters
Vision lets robots identify objects, people, surfaces and motion using compact, widely available sensors.
Advantages
- rich semantic information
- low-cost cameras
- large research ecosystem
Limitations
- lighting and occlusion sensitivity
- training-data bias
- uncertainty can be difficult to calibrate
See it in the Observatory
A definition is a starting point.
Capabilities depend on implementation, operating environment, training data, hardware and safety controls. Follow the connected records and sources before generalizing.