Recognising Traffic Signs

An important skill that a self-driving car must have is the ability to classify traffic signs. In this project it is shown how to perform traffic sign recognition by using Keras and CNNs with Python.

Step 1: Getting setup with Python

For this project, you will need Python 3.5 along with the Tensorflow and Keras libraries.

It is reccomended to install the Anaconda Python 3 distribution from Continuum Analytics.

Choose the appropriate Python 3 Anaconda install package for your operating system here. Download and install the package.

If you already have Anaconda for Python 2 installed, you can create a separate environment for Python 3 and all the appropriate dependencies with the following command:

> conda create --name=yourNewEnvironment python=3.5 anaconda

> source activate yourNewEnvironment

Step 2: Installing Tensorflow

Once you have Anaconda installed, first double check you are in your Python 3 environment:

>python Python 3.5.2 |Anaconda 4.1.1 (x86_64)| (default, Jul 2 2016, 17:52:12) [GCC 4.2.1 Compatible Apple LLVM 4.2 (clang-425.0.28)] on darwin Type "help", "copyright", "credits" or "license" for more information. >>> (Ctrl-d to exit Python)

run the following command at the terminal prompt to get Tensorflow (for Python 3.5, CPU-only):

> pip install --ignore-installed --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-1.2.1-cp35-cp35m-linux_x86_64.whl

then to test if Tensorflow is installed correctly:

> python >>> import tensorflow as tf >>> (Ctrl-d to exit Python)

Step 3: Installing Keras

To install Keras run the following command in the Anaconda environment:

>conda install -c conda-forge keras

and check that the installation worked:

> python >>> import keras >>> (Ctrl-d to exit Python)

Step 4: Download the Data Set (follow instructions in input/download_data.md)

Step 5: Run

Run graphic_test_from_testset.py in order to classify some exemplary traffic signs and see how the algorithm works.