Pose

An object's position plus orientation in space — six degrees of freedom in 3D.

Pose is shorthand for “position plus orientation.” In 3D space, position takes three coordinates — x, y, z — and orientation adds three more rotational degrees of freedom, giving the 6D pose (or 6-DoF pose) that comes up constantly in robotics. A pose is always relative to some coordinate frame — the same cup has different numeric values in the camera frame versus the world frame. Orientation can be stored as a rotation matrix, a quaternion, or Euler angles, and position and rotation are often combined into a single 4×4 homogeneous transformation matrix. A robot typically has to estimate an object's pose before it can grasp it, arm control commonly targets an end-effector pose, and the actions a VLA model outputs are also often an end-effector pose or a delta to one.

ExampleROS's geometry_msgs/Pose message has two parts, position (x, y, z, in meters) and orientation (a quaternion x, y, z, w) — together, exactly one pose.

Also called

6-DoF Pose, 6D Pose

Related

End-Effector Pose · Coordinate Frame · Rotation Matrix · Quaternion · Homogeneous Transformation Matrix · 6D Object Pose Estimation

Sources

Wikipedia: Pose (computer vision)
ROS 2 common_interfaces: geometry_msgs/msg/Pose.msg