Closed nachovizzo closed 2 years ago
Hello, thanks for opening the source code and releasing the data.
We are trying to get the ground truth poses from the arche-flight-8 dataset, and so far we didn't succeed.
We basically created a simple python script which subscribes to the point cloud msg and lookup transformations for the current frame on the tf tree:
import rospy import sensor_msgs import tf def callback(data): timestamp = data.header.stamp trans, rot = listener.lookupTransform("/odom", "/os1_sensor", timestamp) save_pose(trans, rot) if __name__ == "__main__": rospy.init_node("tu_vieja") listener = tf.TransformListener() rospy.Subscriber( "/penguin/os1_cloud_node/points", sensor_msgs.msg.PointCloud2, callback ) rospy.spin()
By inspecting the tf tree we can see that this is how it looks
Is the transformation from odom to os1_sensor the pose obtained by the gt system? If this is not true, how can we obtain it?
odom
os1_sensor
A .txt file witht the poses would also be more than fine for us :)
.txt
Hello, thanks for opening the source code and releasing the data.
We are trying to get the ground truth poses from the arche-flight-8 dataset, and so far we didn't succeed.
We basically created a simple python script which subscribes to the point cloud msg and lookup transformations for the current frame on the tf tree:
By inspecting the tf tree we can see that this is how it looks
Question
Is the transformation from
odom
toos1_sensor
the pose obtained by the gt system? If this is not true, how can we obtain it?A
.txt
file witht the poses would also be more than fine for us :)