Banglish-OCR

Important

You are going to need Tesseract OCR engine to run this app, go to the link below to download and install the engine.

  https://tesseract-ocr.github.io/tessdoc/Downloads.html

Or

  sudo apt install tesseract-ocr

About this Repository

Banglish OCR is an interactive Optical Character Recognition Prototype for Bangla and English. Made with Gradio and Pytesseract.

Clone it

  https://github.com/iamzehan/banglish-ocr.git

Follow the steps below to run this on your local machine.

1. Create Virtual Environment

    virtualenv env
   

2. Activate Virtual Environment

   for bash

    source env/Scripts/activate
   

   for windows

    env/Scripts/activate
   

3. Install requirements

    pip install -r requirements.txt
   

4. run the program

    python app.py
   

5. open your browser and visit the link below while the program is running

    http://127.0.0.1:7860/  
   

6. The app looks like this -

7. End note

runtime.txt controls your python version, use the current version on your machine and install the packages accordingly. Aptfile contains the name of the packages that needs to be installed on your host server's machine. setup.sh contains the port and server name. Procfile runs the command to start the app on the web.