In a previous blog we have shown how complexity may be used in conjunction with streaming video. In this short note we illustrate an example of application of this technology to infer traffic density. The case is that of traffic on a motorway and the video in question is shown at the top of this blog. The video has a duration of 22 seconds and is split into 110 frames. A camera is located above the motorway.
The evolution of complexity – traffic intensity – over time is illustrated below:
A few characteristics points are highlighted below.
The simplicity of this approach to “measuring” traffic intensity stems from the fact that it is not necessary to install sensors and count the actual vehicles. With QCM, we just “look” at the situation and we “feel” the state of the traffic without any effort. This is why QCM is Artificial Intuition.
A final comment: traffic in big cities is highly complex and therefore, in virtue of the Principle of Incompatibility, lacks precision. What the principle asserts is that the more something is complex, the less precise it will be. In other words, one cannot make precise statements about highly complex systems simply because Nature won’t allow it. This, in turn, dictates a more “fuzzy” approach as complexity rises. In the case of traffic is sufficient to say, for example, that is it “normal” or “critical” instead of counting cars and trucks and coming up with precise (but irrelevant) numbers.
Artificial Intuition 
0 comments on “Image Complexity as Proxy of Traffic Density”