In this course, you will learn to build a 3D Object detection system. The course can be found here: https://www.thinkautonomous.ai/point-clouds
The dataset used is the KITTI dataset. I use a subset of 20 point cloud files for this course. Link to the full dataset.
The course is currently being completed by using the dockerfile from this repository here: My Open3D Dockerfile
The bounding boxes are drawn using the point cloud indices that are labelled into similar clusters from the DBSCAN Algorithm using PCA from open3d
