Landmark Detection & Tracking

A small educational project demonstrating robot motion, landmark creation, and sensing in a 2D grid world. The notebooks and helper code show how noisy motion and measurements affect localization and provide the foundation for exploring SLAM (Simultaneous Localization and Mapping) concepts.

Contents

• robot-moving-and-sensing.ipynb — Interactive notebook demonstrating the robot class, movement, landmark creation, sensing, and example data collection.
• landmark-detection-and-tracking.ipynb — (Related notebook) additional experiments and tracking code.
• robot_class.py — standalone robot class implementation
• helpers.py — Utility functions used for visualization in the notebooks

Project overview

The core is a robot class that models:

• Pose (x, y) in a square world
• Motion with Gaussian-like noise
• Measurement/sensing of landmarks with limited range and measurement noise
• Random landmark generation

The notebooks guide you through creating a world, moving the robot, generating landmarks, and collecting measurement + motion data in a time-series format suitable for building or testing SLAM algorithms.

Requirements

• Python 3.8+
• NumPy
• Matplotlib

Install dependencies with pip:

pip install numpy matplotlib

Usage

1. Open robot-moving-and-sensing.ipynb in Jupyter or VS Code and run the cells sequentially.
2. Inspect the robot class, call make_landmarks, move, and sense to collect measurements and data used for SLAM-style exercises.
3. Use display_world from helpers.py to visualize the robot and landmarks.

Notes and next steps

• The code is intended for learning and experimentation rather than production. You can extend it by adding odometry models, probabilistic filtering (particle or Kalman filters), or a complete SLAM backend that ingests the generated data sequence.

Acknowledgments

This project was completed as part of the Udacity Computer Vision Nanodegree Program.