Virtual Calculator using OpenCV

A Python-based virtual calculator that leverages OpenCV for real-time hand gesture recognition, enabling touchless interaction. This project integrates computer vision techniques with a trained Artificial Neural Network (ANN) model to interpret hand gestures as numerical inputs, facilitating arithmetic operations without physical contact.

Features

• Hand Gesture Recognition: Utilizes OpenCV to detect and interpret hand gestures in real-time.
• Digit Classification: Employs a trained ANN model (MNIST_ANN.py) to classify hand-drawn digits accurately.
• Virtual Interface: Provides an on-screen calculator interface that responds to recognized gestures.
• Real-Time Feedback: Displays inputs and results dynamically, enhancing user interaction.

Installation

1. Clone the Repository:

   git clone https://github.com/aksayush2005/VirtualCalculator_using_OpenCV.git
   cd VirtualCalculator_using_OpenCV

2. Set Up a Virtual Environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Dependencies:

   Ensure you have Python 3.x installed. Then, install the required packages:

   pip install opencv-python numpy

Usage

1. Train the ANN Model:

   Before running the calculator, train the digit classification model:

   python CUSTOM_ANN.py

   This script trains an ANN on a custom dataset(custom_hand_data.zip) and saves the model.Ensure the dataset is present in the working directory.

2. Run the Virtual Calculator:

   After training the model, launch the calculator interface:

   python calculator.py

   A window will open using your webcam. Use hand gestures to input numbers and perform calculations.

Project Structure

├── CUSTOM_ANN.py         # Script to train the ANN model on the custom dataset
├── calculator.py        # Main application script for the virtual calculator
├── custom_hand_data.zip # Custom Dataset,ensure that it is present in the working directory

Demo Video:

https://youtu.be/kEPfQGFMVKs

Contributing

Contributions are welcome! If you'd like to enhance functionality, fix bugs, or improve documentation, please fork the repository and submit a pull request.

License

This project is open-source and available under the MIT License.

Acknowledgments

• OpenCV for providing powerful computer vision tools.