Traffic Sign Detection System

A deep learning-based system that can accurately detect and classify traffic signs using Convolutional Neural Networks (CNNs). This project is implemented using Keras and TensorFlow, providing both a trained model and a user-friendly GUI interface for real-time traffic sign classification.

Features

• Traffic sign classification using CNN architecture
• Interactive GUI interface for easy usage
• Support for multiple traffic sign categories
• Pre-trained model included
• High accuracy in classification

Project Structure

• GUI.ipynb: Interactive GUI implementation for the system
• traffic_signs.ipynb: Main notebook containing model training and evaluation
• traffic_classifier.h5/.keras: Trained model files
• Train.csv, Test.csv: Dataset files for training and testing
• Meta.csv: Metadata information
• data/: Directory containing dataset images

Requirements

• Python 3.x
• TensorFlow
• Keras
• NumPy
• Pandas
• OpenCV
• Matplotlib
• Jupyter Notebook

Installation

1. Clone this repository

2. Install the required dependencies:

   pip install tensorflow keras numpy pandas opencv-python matplotlib jupyter

3. Launch Jupyter Notebook to run the GUI:

   jupyter notebook GUI.ipynb

Usage

1. Open GUI.ipynb in Jupyter Notebook
2. Run all cells to start the GUI interface
3. Upload an image of a traffic sign through the interface
4. The system will process the image and display the predicted traffic sign class

Model Information

The system uses a Convolutional Neural Network (CNN) architecture trained on a comprehensive dataset of traffic signs. The model achieves high accuracy in classifying various traffic sign categories.

Results and Screenshots

Epoch-wise Training Output

Training progress log showing accuracy and loss values per epoch in Jupyter Notebook.

Accuracy Plot

Line graph visualizing training vs validation accuracy over training epochs.

Loss Plot

Line graph showing the decrease in training and validation loss over time.

Yield Sign Prediction

GUI displays the predicted label "Yield" after classifying an uploaded image.

Road Work Sign Prediction

GUI successfully identifies and labels the uploaded image as "Road work".

Speed Limit Sign Prediction

GUI correctly classifies the image as "Speed limit (30km/h)".

License

• This project is licensed under the MIT License
• You are free to use, modify, and distribute this software in accordance with the terms of the license