Please use this identifier to cite or link to this item: http://dspace.cas.upm.edu.ph:8080/xmlui/handle/123456789/435
Title: Tess2Speech: An Intelligent Character Recognition-To-Speech Application for Android Using Google's Tesseract Optical Character Recognition Engine
Authors: Baes, Gregorio B.
Custodio, Anter Aaron
Keywords: Android
training
handwritten texts
Tesseract
optical character recognition
computer printed texts
intelligent character recognition
free and non-proprietary
speech-impaired
Issue Date: Jun-2016
Abstract: Tes2Speech is an Android mobile application for recognizing handwritten texts and optionally convert it to speech. By training Google’s Tesseract, which is a free open-source Optical Character Recognition (OCR) Engine, to recognize hand- written texts, I created an alternative for Intelligent Character Recognition (ICR) Engine that are proprietary and expensive. OCR Engines’ main purpose is to con- vert scanned documents with computer printed texts into an editable and machine encoded text. But by improving an OCR into ICR to recognize handwritten texts and integrating it into a mobile application in this day and age, it greatly improves its usability. It can help speech-impaired people to communicate, help children to recognize the proper pronunciation of words they write, etc. Thus, creating a free and non-proprietary handwriting-to-speech application that is accessible for everyone, and will be very useful for those who are looking for a free, portable, and reliable alternative to ICR, OCR, and other related expensive communication devices. Tes2Speech includes a user-friendly Tesseract trainer desktop application in order for the user to personalize Tess2Speech for their own handwritings.
URI: http://dspace.cas.upm.edu.ph:8080/jspui/handle/123456789/435
Appears in Collections:Computer Science SP

Files in This Item:
File Description SizeFormat 
Tess2Speech.pdf10.16 MBAdobe PDFThumbnail
View/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.