Computer vision is different from human vision because computers do not actually “see” images the way humans do. A computer sees an image as pixels and needs algorithms to figure out what those pixels mean. Humans can understand objects, movement, and situations much more naturally, while computer vision usually needs specific conditions and instructions to recognize something.
There are different techniques we can use to help a computer see or track what we are interested in. Some examples are frame differencing, background subtraction, and brightness thresholding. Frame differencing can detect movement by comparing two frames, while background subtraction can separate people or objects from the background. Brightness thresholding can also help detect objects when there is a strong difference in brightness between them and their surroundings.
Computer vision’s ability to track people also connects to surveillance, which can make interactive art more complicated. I think it can be interesting because artists can use tracking to make an artwork react to people, but it also raises questions about privacy and how much people are being watched without realizing it. One artwork I found interesting was Sorting Daemon by David Rokeby. I like it because there is a lot happening in the composite image. The system watches people outside, detects them, separates their images from the background and organizes parts of them by color and size. I think it is interesting because it turns something that feels like surveillance into an artistic image while also making us think about how technology can automatically analyze and categorize people.