Large Vocabulary Arabic Handwriting Recognition System

CFP: #69 Round: 13
Unversity/Faculty: Cairo University Project Type: ARP Company: TAYA IT
ITAC Fund: EGP 936,330 Start Date: 3/4/2014 End Date: Ongoing

Project Output:

Develop A Prototype for an Arabic Handwriting Recognition system that has

  • Reliable Accuracy (> 90%)
  • Large Vocabulary (> 500k words)
  • Real Time Performance (including on mobile devices)

Build a Hidden Markov Model (HMM) based system to provide solutions for most of the difficulties inherent in recognizing Arabic script along with new preprocessing technique that will be applied on the delayed strokes to match the structure of the HMM model.