✍️ Handwritten Digit Recognition (MNIST)

This project uses a Convolutional Neural Network (CNN) trained on the MNIST dataset to recognize handwritten digits (0–9). It includes both a Jupyter notebook for training and testing the model and a Python Tkinter GUI app that allows users to draw digits and see real-time predictions.

---

📦 What's Included

• 🧠 Model Training Notebook – Builds and trains a CNN using TensorFlow/Keras on MNIST
• 🖼️ Interactive GUI (Tkinter) – Lets you draw a digit and predicts it using the trained model
• 📁 Pre-trained Model – Saved as mnist_cnn_model.keras (used by the GUI)

---

🧠 About the Dataset

• MNIST Dataset:
  • 70,000 grayscale images of handwritten digits (28×28)
  • 60,000 training + 10,000 testing samples
  • Classes: Digits 0–9

---

🧰 Tech Stack

• Python
• Jupyter Notebook
• TensorFlow / Keras
• Tkinter (GUI)
• NumPy, Pillow, Matplotlib

---

🏗️ Project Structure

├── mnist_model_training.ipynb (Notebook for model training and evaluation)
├── digit_gui_predictor.py (GUI app to draw and predict digits)
├── mnist_cnn_model.h5 (Saved trained CNN model)
├── requirements.txt (Python dependencies)
└── assets/ ((optional) Sample images or outputs)

---

🚀 Setup Instructions

1. Clone the Repository

git clone https://github.com/Pankaj97825/-Handwritten-Digit-Recognition-MNIST-.git
cd -Handwritten-Digit-Recognition-MNIST-

2.Install Requirements

pip install -r requirements.txt

Or manually:

pip install tensorflow numpy matplotlib pillow

---

✍️ Run the GUI Application

python draw_digit_ui.py

• A window will appear where you can draw a digit with your mouse.

• Click "Predict" to classify it.

• Click "Clear" to draw a new one.

Make sure mnist_cnn_model.h5 is in the same directory!

---

📈 Sample Output

You can expect predictions like:

Prediction: 2 (Confidence: 0.99)

And a GUI like this:

---

📄 License

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

---

🙋‍♂️ Author

Pankaj Kumar
GitHub: @PankajKumar-11