This typology machine will capture recurring forms in architecture around Pittsburgh. I would like to create an animation incorporating photographs of residential and commercial buildings which would be sequenced to create a fluid movement of commonalities through space. I think it’s interesting how we can recognize similar objects as iterations of one another, or iterations of some common master or reference production. This is a property of capitalism that allows for a simulation we take part in where a single idea of a product can be shared with an entire consumer base through its automated reproduction. I’d like to apply this concept to the physical rhythms of architecture, expanding it to a larger and more organized display of repetition.
I was inspired by a similar piece done by video artist Kevin McGloughlin.
In this piece, McGloughlin has taken pieces of these architectural frames (usually kinetic), and composited them into a new collage presenting the patterns in infrastructure as isolated fractals built upon themselves in the new composition.
For my typology I want to only show a series photos I’ve taken rather than compositing something like these. The way that I order and arrange these clips is where the machine comes into play.
I plan on either manually, or automatically finding simple geometric shapes within architectures and labeling a dataset with the combined image and shape data. I would then write a program to find similar shapes across the frames and place these images next to each other in sequence, repositioning, rotating and scaling to match shapes as closely as possible and keep everything in the center of the frame. The result would be a semi fluid animation, rapidly exploring the physical repetitions and commonalities throughout the cities.
