ASCII ARt Photo Booth - 2016

 

For the Secrets of Retail event at Social Media Week 2016 I created a photo booth that turns users’ photos into ASCII art. Users approach the photo booth, pose for 3 photos, and instantly their photos are converted to ASCII art and printed on a low resolution receipt printer (receipt...retail...get it?). A URL on the receipt drives users to a simple web page where they can share their photos and rub off the ASCII characters to reveal the original photos.

Some months before this project, the printer was headed for an Ogilvy dumpster before I rescued it and spent time thinking about ways to celebrate and feature the low res print quality. An early POC is shown in the video at the bottom of the page.

The IBM team later used most of the core features of this project - user flow, screen format, image processing, physical printout, and photo wall - in their IBM Watson photo booth (video).

Role

  • Concept

  • Interaction Design

  • Engineering

Technology

  • Processing with OpenCV

  • Touch screen

  • Camera

  • Amazon S3 storage

  • Star TSP100 receipt printer

The Printed receipt

Here's what the actual printed receipt looks like. The short URL links to web experience below, from where it can be shared on social media. 

 
 

The web experience

The printed receipt drove to this simple web experience in which the user can view and share photos. A little jQuery plugin allows users to rub off the ASCII characters to reveal the original photo.

The printer

This Star TSP100 printer was on its way to the garbage can before I rescued it. Initially I didn't know what I would do with it. Then I realized that everyone loves photo booths, and building one would surely acquire users and give me a chance to experiment with various ways to scale high resolution images down to such a low resolution printer and have it work conceptually.

 
 

A look at an early iteration

While I was building the photo booth, I installed it on a floor in our building to observe usage. Overall I was happy with the user flow of this experience, but it took several iterations to come up with what I thought was the correct image processing technique. In the version of the app seen in this video, I use OpenCV to pick out outlines in the images to simulate a quick pencil sketch, with varying degrees of success.

Earliest proof of concept photo booth app written in Processing. The user approaches the photo booth, takes 3 photos, and receives photos printed on a receipt printer. Image processing attempts to turn the photos into "sketches". Later proof of concept used image processing to turn photos into ASCII art.