The reading shows computer vision as a set of simple tricks for finding motion, color, and change in an image, not as a rich way of seeing like humans have. We can still recognize a friend if the light changes or the background is messy, but a computer vision system often fails unless we control the scene and use tools like frame differencing, background subtraction, and blob tracking. How literal and fragile machine vision really is: it does not “understand” what it sees, it just reacts to pixel changes and thresholds. As designers, we then adjust our projects around these limits, asking people to stand in certain places or wear certain colors so the system can pick them up, which can make the computer seem smarter than it is. The reading also made me realize that even these basic choices what to track, what to ignore already decide which movements and bodies are visible in an artwork and which are not.
The author sounds very hopeful about giving artists and beginners easy tools for using computer vision, and this focus on access feels positive but also a bit one-sided now. The text treats these methods mainly as creative building blocks, but today we know similar techniques are used in security cameras, public-space monitoring, and commercial tracking. Because of that, the cheerful tone around “playful” tracking feels incomplete to me since it does not fully address how normal it can make constant watching feel. The reading did not change my view that computer vision is tied to power and control, but it did sharpen it by showing how quickly simple tools in art and education can be connected to serious surveillance uses. I finished feeling that if we use these techniques in interactive art now, we should make engaging, responsive pieces and try to reveal and question the act of tracking itself.