top of page
BasicAI Wiki Background Image
Wiki

3D cuboid

A 3D cuboid is a 3D bounding box annotation that represents an object’s position, size, and orientation in 3D space. It’s widely used in autonomous driving, robotics, and augmented reality, where depth, volume, and pose matter.

A cuboid is typically defined by its center (x, y, z), dimensions (l, w, h), and rotation angles (yaw, pitch, roll). Data annotators create 3D cuboids in LiDAR point clouds or in projection views that combine camera and sensor data.

Compared with 2D bounding boxes, 3D cuboids provide richer spatial information and are a standard label type in 3D detection benchmarks such as KITTI and nuScenes.

LiDAR

3D cuboid annotation in LiDAR point cloud

Related

Term

LiDAR

Term

Point Cloud

E-Book

3D LiDAR Point Cloud Annotation Guide

CTABack-min.png

Transform your vision data into AI training sets with unmatched precision.

Contact our team for a consultation and custom quote tailored to your specific project requirements.

bottom of page