EC-thor leverages Euclidean clustering to identify distinct groups of points in a points buffer, enabling efficient cone detection. The algorithm is tailored for the unique challenges of the autonomous system of a Formula Student car, specifically developed for BCNeMotorsport team.

Pipeline diagram

• FEC: Fast Euclidean Clustering for Point Cloud Segmentation
• ikd-Tree: Incremental KD-Tree for Robotic Applications

• Adaptation to Point Net

• Modification Acumulator.hpp

• Parallelization of some functions