Obsessive Character Recognition is a comment on the fallibility of algorithms. These tools we rely on so much in our increasingly digitally mediated lives are often considered neutral and incorruptible but, in reality, are black boxes susceptible to the biases of their creators.
Obsessive Character Recognition uses Optical character recognition algorithms to explore this, by requiring the algorithm (in this case we use one called Tesseract) to find text in a video stream from an environment where there may be none it highlights both the restrictive nature of algorithmic tools and their ability to reshape our world by the way they analyse, process and represent it.
The project takes the form of a piece of software that uses a webcam to gather the environment and presents the text mapped loosely to its structure, moving the characters in 3d to reform the space from text. There are a number of user adjustable presentational and functional parameters.
[The code is not really ready for public consumption, as yet, and has a few awkward dependencies but if you’re interested, let me know and I am happy to send it to you]