Page Comparison

There is a list of targets that Nao can track using the Naoqi Tracker API, which works well under good lighting conditions. Here is a walkthrough on how to track a face with its head and a red ball with its arms.

Track a face with Nao’s head

1. First, import the necessary functions from the SIC framework:

...

4. Since we want to use the head to track an objecta face, we need to set the stiffness to 1 to enable the motor motors to control the head joint

...

nao.tracker.request(
    StartTrackRequest(target_name=target_name, size=0.2, mode="Head", effector="None")
)

6. Unregister a target

nao.tracker.request(RemoveTargetRequest(target_name))

Track a red ball with Nao’s arms

1. Set a new target

target_name = "RedBall"

2. Set a robot position relative to target. This is only needed in Move mode. The parameters represent [coordX, coordY, coordWz, thresholdX, thresholdY, thresholdWz]. In this case, the robot stays 30 cm away (along the x-axis) with a 10 cm threshold from the red ball. The XYZ frame reference for the robot can be found here position-identification

move_rel_position = [-0.3, 0.0, 0.0, 0.1, 0.1, 0.1]

3. Track a red ball

nao.tracker.request(
    StartTrackRequest(
        target_name=target_name,
        size=0.06,
        mode="Move",
        effector="Arms",
        move_rel_position=move_rel_position,
    )
)

4. Stop all tracking. This will put the robot into rest mode, unlocking all joints and setting stiffness to 0

nao.tracker.request(StopAllTrackRequest())

A complete script for this tutorial can be found here: demo_nao_motiontracker.py