Computer vision works in a very different way from how people see. We read situations without thinking. We notice feelings, intent, and small details that are not written anywhere. A computer has none of this. It only receives a grid of values and waits for someone to tell it what to look for. Nothing has meaning until a developer builds rules or trains a model to reach a goal.
There are simple methods that can guide the system. Finding outlines helps the computer separate one shape from another. Tracking color ranges can point it toward a moving object. Contours make it easier to follow form. With the right setup and some training, the computer can get close to noticing what matters in a scene, even if it never truly understands it.
This matters a lot in interactive art. Vision gives the work a way to notice the audience and respond. A camera can capture movement, distance, and presence at the same time, which makes it one of the strongest sensors an artist can use. At the same time, people take in art through many senses. We see, listen, and feel. Computer vision tries to narrow that gap by giving machines a way to observe instead of just wait. This opens new creative space, but it also raises questions about how much the artwork should watch and how that changes the experience for the viewer.