What A Physics Student Can Teach Us About How Visitors Walk Through A Museum

Unfortunately the article appears to be a lot of conjecture based on about 20 minutes of manual observations. While it's interesting, there's a lot more work required before you could get close to drawing conclusions.

I think a really interesting thing to do would be to place some cameras overhead, facing down, and then use either image recognition techniques (eg OpenCV), infra red cameras and heat tracking, or even something like a kinect to track the points at about head height. You could build data up over days or weeks rather than minutes, and get a much better idea of how people move through the space.

Could this be used to further analyse websites? Instead of just the usual eye heat map, and optimising for that would it be better to optimise for each page based on the content on that page?