Download presentation
Presentation is loading. Please wait.
1
iRead Team Members: Matt Cardin, Paul Kennedy, Oscar Perez
2
Why The iPhone Wanted to work with mobile devices Its available to us iPhone development is well supported Objective C looks challenging
3
Can't Read? There's an app for that iRead is an iPhone application It really reads text!
4
How to read Open the app Hold the iPhone camera up to some text Hit the read button Listen to your iPhone read the text to you Its that easy!
5
How it reads Camera takes picture OCR gets the text Text-to-speech speaks the text
6
OCR (Optical Character Recognition) Input picture, output text Current candidate: Tesseract (written in C and C++)
7
Character to speech engine Input text, output audio Current candidates: FLITE and eSpeak (both written in C)
8
This project seems simple Developing OCR and text-to-speech is NOT the goal Project Challenge: to port open source projects to the iphone, and glue OCR with text-to- speech efficiently Other Goals: Learn a new language (objective C), experiment with cross compilation, build a neat UI
9
iRead (full version) Now with... Translation Long sentence scanning
10
Translation Tesseract can handle any Unicode character eSpeak designed for multiple languages Would keep translation control in the user interface
11
Scanning Will be able to capture longer sentences that don't fit in one image Multiple lines?
12
Scanning, is it necessary? Tesseract: font x-height min = 20px Camera quality text means 40px Sqrt of 2 Megapixels = 1448x1448 pixels 1448 / 40 = 36 characters = about 5 words So yes, we could use scanning
13
Work Division Matt – text to speech Paul – OCR Oscar – low level porting and cross compilation
14
Questions?
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.