The "Currency Detection for the Visually Impaired" project aims to assist visually impaired individuals in identifying different banknotes accurately. The system leverages Convolutional Neural Networks (CNN) to recognize and classify various currencies in real-time through a live camera feed.
-
Firstly clone the repository.
git clone https://github.com/OMKAR-KALEPU/the-blind-assist.git
-
Download the Dataset.
-
Rename the denomination folders to:
1Hundrednote
2Hundrednote
2Thousandnote
5Hundrednote
Fiftynote
Tennote
Twentynote
-
Create a virtual environment.
python -m venv .venv
-
Activate the virtual environment. Open the terminal or command prompt and give the following command.
.venv\Scripts\Activate
-
Now that the virtual environment is activated, install the necessary libraries as specified in requirements.txt
pip install -r requirements.txt
-
You can create you own classifier by executing the models
(OR) You can feel free to download the already trained DenseNet Classifier here. -
Now you can start the django server.
python manage.py runserver
-
Now the project is live at
http://127.0.0.1:8000/
-
The user interface looks something like this:
- You can click on the User section present in the navbar and register yourself and then you'll be able to predict your notes using live camera with voice output.
-
The path to the classifier should be changed accordingly in user/views.py file.
-
The integrated web cam might not be able to detect the note properly if the quality of camera is not good. So I recommend using an external cam to be added to it. If external cam is added, the a line of code in user/views.py should be modified to:
cv.videoCapture(1)
Thank you :)