Autonomous Navigation & Object Localization in ROS2 (Course Project)

Task 1 — Autonomous Mapping

Key Components

• LiDAR-based perception: Reads /scan data to detect obstacles, track wall geometry, and determine free space.
• FSM architecture: Modular states for exploration, turning, and wall-following, allowing context-aware navigation.
• Control system:
  • PI controller for heading stabilization
  • Constant-speed forward motion for smooth exploration

Demo

Task 2 — Navigation with Static Obstacles

Key Components

Implemented a two-level navigation stack combining A* global planning with RRT local replanning from real-time /scan data.

Demo

Task 3 — Search and Object Localization

Key Components

• Built a vision–LiDAR fusion pipeline for object localization:
  • Color-based contour detection
  • Image-angle computation
  • LiDAR distance estimation
  • World-coordinate triangulation with RViz visualization
• Integrated a waypoint-based patrol strategy to search the entire map and reliably detect targets.

Demo