BEV Projection for SemanticKitti Dataset

This repository is for multi-channel Bird-Eye-Projection for SemanticKitti dataset. This package produces rgb images using the point cloud data and semantic labels, where each color represents a specific semantic class.

Requirements

• PCL
• OpenCV

Data Structure

Please download SemKITTI dataset and ensure the following structure

./dataset/
├── 
└── SemanticKitti/
    ├──sequences
        ├── 00/           
        │   ├── velodyne/	
        |   |	├── 000000.bin
        |   |	├── 000001.bin
        |   |	└── ...
        │   └── labels/ 
        |   |   ├── 000000.label
        |   |   ├── 000001.label
        |   |   └── ...
        |   └── BEV_images/ 

Run Packages

./BEV_projection /home/user/dataset/SemanticKitti/sequences/00/

Example Output