lidar_point_clouds

Introduction

Each of the folders here correspond to the various tasks of our project.

The visualize folder holds the Python files for visualizing in Open3D and PPTK.

The fuse folder holds the code for performing sensor fusion as well as the MATLAB developer kit code from the KITTI website. The original code can be found here.

The segmentation folder holds the code for filtering via segmentation.

The clustering folder holds the code for performing various clustering methods on the point clouds.

The bboxes folder holds the code for filtering via bounding boxes.

The misc folder holds various Bash scripts for modifying the data formats or extracing a sample of the data.

The data folder is ignored in the .gitignore. The data_loader.ipynb file has the commands to download the data from the KITTI website.

Instructions

Create a Conda virtual environment with conda create -n lidar python=3.6. Open3D and PPTK only work on Python 3.6.

Run mkdir data to create the data folder.