Computer vision
컴퓨터 비전
Also known as: vision AI
The field concerned with what is in an image or video, and where.
It splits into classification (what is this), detection (where is it), segmentation (which pixels) and tracking (how it moves across frames).
In early education it links objects in a photo to words. For only the tapped object to light up, segmentation is needed rather than detection.
It is sensitive to lighting, angle and occlusion, so how a product behaves when the result is uncertain decides its quality.
- LadderClassification, then detection, then segmentation.
- User guidanceTelling the user to retake the photo matters as much as the model.