Autonomous Driving - Lane and Car Detection

Description

The aim of this project was to develop an application, that from a video and making use of the GPU for image processing, detects lanes and vehicles. Furthermore, it marks the car depending on the lane they are detected:

1. Draws a red square, in case the other vehicles are in the same lane.
2. Draws a green square, in case they are detected in other lanes (see the picture).

How it works

In case you want to explore and improve the code, the application works as the following:

1. Install pycharm (recommended) or any other IDE that can run python code.
2. Install the required packages through the pip3 install command. You should only need to install the OpenCv library and OpenCl library:
   • pip3 install opencv-python
   • pip3 install pyopencl
3. Run the code using the play button of the pycharm IDE (in case you are using pycharm).

The video should start playing and you should see:

1. Two lines (blue and red) detecting the lanes.
2. A square (red and or green) identifying the cars.

Implementation

The system was implemented according to the following diagram:

Questions or Future Improvements

If you would like to make improvements or you have any other questions, you can contact me here on GitHub or on LinkedIn which link is on my GitHub profile page. I would be glad to answer any comments or questions.