Build a Traffic Sign Classifier

Overview

In this project, I used Python and TensorFlow to classify traffic signs.

Pipeline architecture:

• Load The Data
• Dataset Summary & Exploration
• Data Preprocessing
  • Shuffling
  • Grayscaling
  • Local Histogram Equalization
• Normalization
• Design a Model Architecture
  • LeNet-5
  • AlexNet
• Model Training and Evaluation
• Testing the Model Using the Test Set
• Testing the Model on New Images

Dependencies

This project requires Python 3.5 and the following Python libraries installed:

• Jupyter
• Numpy
• Matplotlib
• OpenCV
• Scikit-learn, Scikit-image
• Pandas
• Tensorflow

Dataset

Dataset used: German Traffic Sign Dataset. This dataset has more than 50,000 images of 43 classes. You can download the dataset from here. We already have three .p files of 32x32 resized images:

• train.p: The training set.
• test.p: The testing set.
• valid.p: The validation set.

Preferences

CarND-Traffic Sign Classifier Project - Udacity