ROS2 Self Driving Car with Deep Learning and Computer Vision

Develop a ROS2 self-driving car from the ground up using an RGB camera! This comprehensive course guides you through building a functional autonomous vehicle, covering key self-driving car features and practical implementation.

Learn to implement:

• Lane Assist: Keep your car safely within its lane.
• Cruise Control: Maintain a set speed for comfortable driving.
• T-Junction Navigation: Successfully navigate complex T-intersections.
• Crossing Intersections: Safely handle crossing intersections with intelligent decision-making.

Course Content:

• ROS2 Fundamentals: Master ROS2 concepts, including nodes, launch files, and world models.
• Gazebo Simulation: Create and edit world models and car parts using Gazebo. Explore SDF (Scene Description Format) for advanced scene creation, including textures and plugins.
• Computer Vision for Autonomous Driving: Implement lane detection using advanced computer vision techniques.
• Object Recognition: Employ sign classification with custom-built CNNs to identify road signs.
• Traffic Management: Develop traffic light detection using Haar Cascades.
• Tracking & Control: Implement sign and traffic light tracking using optical flow. Learn rule-based control algorithms for autonomous navigation.

Software Requirements:

• Ubuntu 20.04 (LTS)
• ROS2 - Foxy Fitzroy
• Python 3.6
• OpenCV 4.2
• TensorFlow 2.14

Course Flow (Self-Driving Development Stages):

We'll start with a rapid deployment of your car on a Raspberry Pi using pre-built 3D models (provided) and affordable car components (links provided). Then, we'll move to serious programming by interfacing the Raspberry Pi with motors and the camera.

Gain insights into the future of transportation and the environment through a comparison of leading autonomous vehicle companies like Tesla and Waymo. We'll present our project directly within a simulation, allowing you to witness the course outcomes firsthand.

Key Features Development (Each with Detection & Control components):

1. Lane Assist
2. Cruise Control
3. T-Junction Navigation
4. Crossing Intersection

Prerequisites:

• Basic ROS2 node communication understanding.
• Fundamental CV (Computer Vision) knowledge.
• Experience with launch files is helpful.
• Familiarity with Gazebo model creation.
• A motivated mind and passion for programming!

Ready to dive in? Explore our GitHub repository or contact us to access the code and begin learning! (Even if you don't purchase the course, accessing the code is highly recommended for learning.)