Monocular SLAM Using Ground Textures

We implemented and enhanced the monocular ground-texture-based SLAM algorithm described by Hart et al. (ICRA 2023). Key components: - Using ORB for detecting and describing keypoints, with experiments using SIFT, SURF, and HARRIS for better accuracy. - FLANN for efficient keypoint matching. - Visual Bag of Words to identify revisited locations and improve loop closure detection. - Factor Graph Optimization with M-Estimators to refine pose estimates and handle outliers. We improved the original algorithm by introducing additional ground textures and testing alternative keypoint descriptors. After validating with public datasets, we collected and tested data using a custom TurtleBot setup, assessing the algorithm in scenarios with changing textures, uneven surfaces, and varying lighting.