“…Christensen says they have been able to identify unique characteristics of buildings that point to stylistic influences beyond any initial blueprints.”
An art history academic is using computer vision to gather new insights into the creative forces behind architecture.
Writing for The Conversation, Peter Christensen, an Assistant Professor of Art History at the University of Rochester, says that he was inspired by the emergence of facial recognition technology. One particular inciting incident got him going, when the face-tagging feature of his iPhone confused one of his friends with the famous Great Mosque of Cordoba.
As he tells it, Christensen began to realize that this kind of computer vision technology could be used to look at buildings in new ways. He started “[t]hinking about buildings as objects with biometric identities,” and worked with a research team to build 3D scanning systems that could create digital mockups of buildings that could then be analyzed in a process “similar to facial recognition”.
In this way, Christensen says they have been able to identify unique characteristics of buildings that point to stylistic influences beyond any initial blueprints. Looking at railway stations built by foreign workers from a range of backgrounds in Canada in the late 19th century, for example, the researchers can use this technology to discover beveling on window frames or pointed arches that offer hints about who was working on these particular aspects of a building.
It’s a particularly niche example of how computer vision technologies have evolved to help people process visual data in new ways – to see the world, in other words. And while some facial recognition technology remains lamentably inaccurate in matching human faces, at least such inaccuracies can offer some surprising new insights in other applications.
Source: The Conversation (via CNN)
July 26, 2018 – by Alex Perala