🖍️ Handwritten Digit Recognition System

This project allows users to recognize handwritten digits using a deep learning model (CNN) trained on the MNIST dataset.
Users can either:
✔️ Upload an image of a digit
✔️ Draw on a writing pad and get instant predictions

---

✨ Features

✅ Image Upload – Choose an image of a digit and get a prediction
✅ Interactive Writing Pad – Draw a digit using a mouse
✅ Clear Canvas (Press 'C') – Erase and redraw without restarting
✅ Predict Digit (Press 'P') – Get the result instantly
✅ Quit (Press 'Q') – Exit the writing pad

---

🛠️ Installation

1. Clone the repository:

   git clone https://github.com/your-username/handwritten-digit-recognition.git
   cd handwritten-digit-recognition

2. Install dependencies:

   pip install opencv-python numpy tensorflow tkinter

3. Download the trained model:

   • Place the pre-trained handwriting_model.h5 inside the models/ directory.

---

🚀 Usage

Run the program using:

python main.py

You’ll see a menu asking you to:
1️⃣ Upload an image
2️⃣ Draw on a writing pad

For Image Upload:

• Select an image (.png, .jpg, .jpeg) containing a digit.
• The system will process and display the predicted digit.

For Writing Pad:

• Draw a digit using your mouse.
• Press 'C' to clear and redraw.
• Press 'P' to predict the drawn digit.
• Press 'Q' to exit.

---

📂 Project Structure

📂 handwritten-digit-recognition
│── 📂 models
│   └── handwriting_model.h5       # Pre-trained CNN model
│── digit_recognition.py           # Main program to Train & Run the Model
│── realtime_digit_recognition.py  # Program to Acces the Local File & Writting Pad
│── README.md                      # Project documentation

---

🧠 How It Works

• Image Processing: Converts images to grayscale, resizes (28x28), normalizes, and inverts colors.
• CNN Model Prediction: Trained on MNIST dataset, predicts digit with softmax classification.
• OpenCV GUI: Provides an interactive writing pad for drawing digits.

---

🖥️ Demo

📸 Screenshot 2: Writing Pad in Action

📸 Screenshot 3: Image Upload Prediction

---

🤝 Contributing

Want to improve this project? Feel free to fork and submit a pull request!

---

📜 License

This project is open-source and free to use! 🚀