3D bounding box visualization using open3D

This repository is the visualization of 3D Bounding Box in Point Clouds using open3D library.

• Install open3D Library

pip install open3d-python

• Please first run view_point_experiment.py to adjust your view point based on your monitor and preference. This file will generate a json file to represent the view parameter, please replace the json file in video_generator.py.

• In order to direct use this code, please change your label file to the same format of KITTI dataset. Or modifiy the function in helper_functions.py and vis_utils.py.

• Run vis_train_val.py to get your required output. Need to set the correct path in the code.

1. image_with_gt_2d_bbox： Draw 2d ground truth bounding box into images
2. image_with_3d_bbox : Draw 3d bounding box into images
3. image_for_point_cloud_wo_bbox : Raw point cloud images
4. image_for_lidar_with_3d_bbox : Draw bounding box into 2d bird eye view images
5. image_for_point_cloud : Point Cloud with 3d bounding box

• Parameter.py contains all the necessary parameters when generating output images.

• Example output video please see: https://youtu.be/7IUUmFAFa-A