← All explainers

Computer vision analyses images

Updated: 0 views

Computer vision extracts features and information from photographs and video.

Different tasks

Classification labels an image; detection locates objects; segmentation marks regions. One task’s accuracy does not describe another.

Test real conditions

FinFly recommends using your lighting, angles and cameras, not only selected demonstration images.

Measure error costs

Separate missed objects from false detections. Add human review for consequential decisions.

Primary sources

The explainer is checked against these institutional and industry sources. Links open the original material.

All explainers