Autonomous_Vehicles_Specialization

Projects while studying the Autonomous Vehicles special course by University of Toronto

Course 1 : Mechanical Modeling & Control Methodologies

• kinematic, dynamic modeling of the autonomous vehicle model
• system modeling of the environment: rule-based, real-time
• error dynamics and optimization
• longitudinal, lateral controllers (Pure pursuit, Stanley, MPC algorithm)

Course 2: Localization

• state estimation using Least Squares method and KalmanFilter Algorithm
• sample question for extended KF algorithm
• state estimation and path finding using ErrorState-ExtendedKalmanFilter(ESEKF) Algorithm with SensorFusion (LiDAR, GNSS, IMU) + Visualization

Course 3: Perception

• 3D Camera Perception via geometry and disparity algorithm
• Estimate drivable plane with semantic segmentation and RANSAC algorithm
• Lane estimation with Canny Edge Detection and Hough Tranformation algorithm with threshold filtering & merging
• Computing minimum distance for collision avoidance with object detection and semantic segmentation

Course 4: Motion Planning

• Mapping the gathered sensor data of external environment into proper map type in real-time (e.g. Ocupancy Grid)
• Find the most efficient route with search algorithm
• Total Motion Planning:
  • Local Planner
  • Collision Checker
  • Behaviour Planner
  • Path Optimization
  • Velocity Planner