Blind-Vision

Created an application for visually impaired people to recognize the Indian currency with their mobile.

Developed an android application. Created a deep learning YOLOv3 object detection model to detect currency notes and served it with a REST API created with django.

Demo:

See the demo video here: here

REQUIREMENTS:

django==3.1.4

opencv-contrib-python==4.4.0.46

numpy==1.18.5

django-rest-framework

also given in the requirements.txt file.

1. Intall the python libraries and start the django server

pip install -r requirements.txt

cd currency

python manage.py runserver

This command will run your localhost server will give you a link as -

http://127.0.0.1:8000

This means your application is hosted on port number 8000
Leave you terminal open and let's go further

To use this API in you mobile application you have to host it somewhere, so that you can call the API from other devices.
To turn my localhost server to live, I use ngrok
Download it from the above link and open. A command prompt/terminal will appear. Now execute the below command there.

ngrok http 8000

This will create a link for your live application.
Leave the command prompt/terminal open

2. Intall the flutter application

Make you have connected an AVD or you phone usb debugging
Excute the following commmands

cd customCamera

flutter pub get

Open the file customCamera/lib/main.dart Copy the http link from ngrok and paste it in the main.dart at line 48
Make sure you don't remove the path in the domain. That is the actual endpoint of the api.

Now make sure you django localhost server and ngrok is running
After pasting the URL, execute the below command in the cmd/terminal -

flutter run

This will install the app on your mobile and you can test it.