By: Hadley Scholtz Supervisor: Mehrdad Ghaziasgar Co – supervisor: James Connan Assisted by: Ibraheem Frieslaar
Introduction User Requirements Requirements Analysis Assumptions Project Plan
Problem ◦ Users with impaired vision or users that are illiterate will have problems reading certain segments of text. ◦ Text in foreign languages are difficult to read, pronounce etc.
Phone Reader 1.0 by Hadi and Hossein Shayesteh ◦ An excellent novel idea by James Connan ◦ User would use Android phone to take picture of text ◦ Picture was sent to Optical Character Recognition (OCR) server ◦ Image converted to text and possibly translated. ◦ Text read to user using text-to-speech (TTS) Features ◦ Could process image, if well-aligned ◦ Entire image was processed and converted to text ◦ Audio was played immediately after processing Previous Solution
Phone Reader 2.0 ◦ Optimized OCR ◦ Select segments from image ◦ Manually control when audio is played ◦ Repeated audio playback ◦ Enhanced user interface
Android Phone Text
Android Phone OCR Server Tomorrow OCRTTS
Process one image at a time Will run on Android platform
Alginahi, Y. (n.d.). Preprocessing Techniques in Character Recognition. Bradsky, G., & Kaehler, A. (2008). Learning OpenCV Computer Vision with the OpenCV Library. California: O'Reilly Media Inc. Seeger. (2003). Patent No. US B1. United States of America. Seeger, M., & Dance, C. (n.d.). Binarising Camera Images for OCR.