While we had completely the majority of the project in the weeks prior, there were still a few bugs that needed to be ironed out and artistic details that needed to be added before we could officially call the project done.

One of the main issues that we ran into was the fact that the camera would not pick up the entire area the people interacting with our project would be stationed in. We took the camera and moved it up until the full area was in view, which ended up with us elevating the camera around 2ft in length. However, now the light would reflect on the vents within the camera’s vision. As a result, we would get some faulty results in our program and thus, the LED lights would shine in the wrong locations. To correct this, we had to reposition the camera in the newly elevated position so no vents or reflective surfaces were in view. After making this change, the computer vision correctly picks up objects within the view.

Another challenge we had to face was how to make those within the installation visible to the camera. From the get-go, we knew that we would have to make some sort of hat. We initially just tried using normal hats which were all varying colors. This approach did not provide enough of a colorful surface for the camera to accurately pick up. As a result, we attached sheets of construction paper to the top of the hats. While ridiculously stupid looking, it proved to be an effective way to make the camera detect different entities.

While a little silly looking, the overall additions were effective and the installation works as envisioned. Before the final art show itself, we might add a few small details, but the overall installation is pretty much complete.