eScanor

electronic scan optical recognition

(the c in character isn't visible because you can't c it :p )

overview

University final project for the curriculum subject named 'Interfacing 1'. Introducing briefly, eScanor allows an impaired vision individual to scan and read text from an A4 paper. It takes a snapshot of the image, converts to a txt file using OCR (Optical Character Recognition) and then speaks that converted text using TTS (Text-To-Speech). It is designed to be as simple and minimal with minimum hassle for interaction, just a single push of a button will automatically trigger the procedure of scanning,convert and speak.

limitations

• Using the 5MP camera that were built for the RPi and it isn't monochrome. Therefore, many noises may come from the surroundings.
• Lighting and environment plays a crucial part of taking image and converting to plain text.
• Takes significant amount of time to scan and convert for lengthy worded paper.

Utilities

• tts.sh -- Tests the speaker for tts and adjust different voices.
• test.sh -- Tests each component including camera.

init

1. sudo apt-get update
2. sudo apt-get install -y build-essential python-pip python-dev
3. sudo apt-get install -y tesseract-ocr flite

optional

4. crontab ./cronfile

testing the camera

head to raspberrypi.org documentation site