Hi @WujiaPan,
Here is sample code that converts the objects’ positions into sl.Pose objects:
# Enable object detection
zed.enable_object_detection()
objects = sl.Objects()
...
# Inside grab() loop
...
zed.retrieve_objects(objects)
for obj in objects.object_list:
obj_translation = sl.Translation()
obj_translation.init_vector(obj.position[0], obj.position[1], obj.position[2])
obj_transform = sl.Transform()
obj_transform.set_translation(obj_translation)
obj_pose = sl.Pose()
obj_pose.init_transform(obj_transform)
# Use the camera_to_geo method here