✍️ Handwritten Digit Recognition (CNN)
A deep learning project that recognizes handwritten digits (0–9) using a Convolutional Neural Network (CNN) trained on the MNIST dataset. Built using TensorFlow and Python, this project demonstrates image classification with high accuracy and real-time prediction capability.
🚀 Features
- 🔢 Recognizes handwritten digits (0–9)
- 🧠 Built using Convolutional Neural Networks (CNN)
- 📊 Trained on MNIST dataset
- ⚡ Real-time prediction support
- 📈 High accuracy on test data
- Python
- TensorFlow / Keras
- NumPy
- Matplotlib
- OpenCV
📂 Dataset
This project uses the MNIST Handwritten Digits Dataset, which contains:
- 70,000 grayscale images (28x28 pixels)
- Digits from 0 to 9
📌The dataset is automatically loaded using TensorFlow:
from tensorflow.keras.datasets import mnist⚙️ Installation & Setup
- Clone the repository:
git clone https://github.com/YOUR_USERNAME/REPO_NAME.git
cd REPO_NAME- Install dependencies:
pip install numpy pandas tensorflow matplotlib opencv-python- Run the notebook:
Open in Google Colab
- Train the CNN model on MNIST dataset
- Evaluate model performance
- Test predictions on sample images
- (Optional) Use real-time input for digit recognition
📊 Results
- Achieves high accuracy on test dataset
- Efficient classification of handwritten digits
📌 Future Improvements
- Improve model accuracy with deeper architectures
- Add GUI for user input
- Deploy as a web app
This project is licensed under the MIT License.