Peeling Off The Painted Layers of NYC Walls: Experiments With The Google Street View Archive

By Brian Foo, NYPL Labs
June 24, 2014
Stephen A. Schwarzman Building

As a web developer who works on a screen and an illustrator that works on paper, I have always admired those who could paint big—often on impossibly large and inconveniently placed walls—only to be erased in a matter of weeks or days. The ephemeral nature of street art is what makes it simultaneously appealing and frustrating as a viewer. However, Google Maps recently rolled out a feature allowing users to go back in time on its Street View. I immediately thought to check out the well-known wall on Bowery & Houston and found that Google captured the painted wall dating back to 2007. Here's a sampling from 2007 to present. I added a few images of the wall that I found while perusing the web to fill in some of the gap years that Google didn't capture.

Next, since the images of the walls were taken from different angles, I built a very basic web tool to align them by defining each of their four corners to be used as control points.

Once all the control points are defined, the tool then warps each image so they look like that are shot at the same angle.  Below, you can see before and after the images were warped:

Next, since all the walls were aligned, I thought it would be interesting to lay them on top of each other chronologically, where the most recent wall is on top and the oldest wall is on the bottom. I created a brush tool that allows you to "erase" each layer to reveal the one previous to it.

You could also cut holes on each layer to get disorienting results like this:

Or what if we put walls in reverse order, so you are painting on top of existing art.  Here's an example using the tool with 5 Pointz in Long Island City before and after it was whitewashed:

Or why limit ourselves to street art? There are many other things that would be interesting when organized in chronological layers:

As more and more digital materials become available from the library and beyond, we will need to continue to ask what we should be remembering and what tools we can build to surface topics of interest and encourage conversations around them.

All the code for this tool is publically available here.