6533b7d7fe1ef96bd12680a6

RESEARCH PRODUCT

Iris : a solution for executing handwritten code

subject

description

Masteroppgave i informasjons- og kommunikasjonsteknologi IKT590 2012 – Universitetet i Agder, Grimstad This paper presents a novel approach to executing handwritten code, the solution coined Iris. My research falls within the field of mobile app development, handwriting recognition, optical and intelligent character recognition (OCR & ICR), machine learning, as well as various Computer Science-related fields such as domain specific languages, or DSLs. The solution outlined in this paper details a system where one can author code using only a writing utensil (such as a pen), scratch paper (such as a napkin), and a smart phone. Iris leverages the power of the cloud to process an image of handwritten code and return the result to the user. Ultimately, my results show that Iris was able to accurately execute handwritten scripts with various levels of observed accuracy. Future work includes adding more layers of machine learning as well as further pre-processing images prior to OCR.