I am subscribing to topic "/camera/depth/points"
and message PointCloud2
on a turtlebot (deep learning version) with ASUS Xtion PRO LIVE camera.
I have used the python script below under the gazebo simulator environment and i can receive x, y, z and rgb values successfully.
However, when i run it in the robot, the rgb values are missing.
Is this a problem of my turtlebot version, or camera or is it that i have to specify somewhere that i want to receive PointCloud2 type="XYZRGB"
? or is it a sync problem? Any clues please thanks!
#!/usr/bin/env python
import rospy
import struct
import ctypes
import sensor_msgs.point_cloud2 as pc2
from sensor_msgs.msg import PointCloud2
file = open('workfile.txt', 'w')
def callback(msg):
data_out = pc2.read_points(msg, skip_nans=True)
loop = True
while loop:
try:
int_data = next(data_out)
s = struct.pack('>f' ,int_data[3])
i = struct.unpack('>l',s)[0]
pack = ctypes.c_uint32(i).value
r = (pack & 0x00FF0000)>> 16
g = (pack & 0x0000FF00)>> 8
b = (pack & 0x000000FF)
file.write(str(int_data[0])+","+str(int_data[1])+","+str(int_data[2])+","+str(r)+","+str(g)+","+str(b)+"\n")
except Exception as e:
rospy.loginfo(e.message)
loop = False
file.flush
file.close
def listener():
rospy.init_node('writeCloudsToFile', anonymous=True)
rospy.Subscriber("/camera/depth/points", PointCloud2, callback)
rospy.spin()
if __name__ == '__main__':
listener()
depth-registered
as well but i cant remember now, i will check this once – Burdensomedepth_registration:=true
so i will try this tomorrow morning. 1 question, is using openni for this the most normal approach? – Burdensome/camera/depth/points
only shows xyz, whereas/camera/depth_registered/points
shows the combination of xyz with rgb. Funny thing is why is it different in gazebo.. well, never mind. If you post it as an answer i will accept it. – Burdensome