How to make the subscribe node to receive "/camera/depth/image_rect_raw"?

[kevinDrawn]

Hello. I have question about the topic named '/camera/depth/image_rect_raw'
This topic is published when i launch this 'rs_camera.launch'.

I want to receive the topic and transfer to depth_frame for mapping x,y pixel coordinate to the world coordinate (x,y,z).

ex) The mapping function is it. (Depth -> Relative Coordinate). This functon's parameter requires the depth_frame.
def get_relative_xyz(depth_frame, pixel_x, pixel_y, depth_scale=10): depth_intrin = depth_frame.profile.as_video_stream_profile().intrinsics depth_pixel = [pixel_x, pixel_y] depth_image_point = rs.rs2_deproject_pixel_to_point(depth_intrin, depth_pixel, depth_frame.get_distance(pixel_x, pixel_y)) depth_relative_point = np.array([depth_image_point[0]*depth_scale, depth_image_point[1]*depth_scale, depth_image_point[2]*depth_scale]) return depth_relative_point

And This is my code.

`#!/usr/bin/env python
#-- coding:utf-8 --

import rospy
from sensor_msgs.msg import Image
from cv_bridge import CvBridge
import cv2
import pyrealsense2 as rs
import numpy as np

x = 250
y = 250
i = 0
depth_relative_point=0

bridge = CvBridge()

pipeline = rs.pipeline()
serial_number = "128422271045" # it is the Serial number of my Depth camera.
config = rs.config()
config.enable_device(serial_number)
config.enable_stream(rs.stream.depth, 848, 480, rs.format.z16, 30) #launch file
pipeline.start(config)

def Image_callback(image):
global x, y, depth_relative_point
depth_array = np.array(image.data, dtype=np.uint16)
depth_frame = rs.frame(depth_array)
depth_relative_point = get_relative_xyz(depth_frame,x,y,depth_scale=10)
print(depth_relative_point[0])

def get_relative_xyz(depth_frame, pixel_x, pixel_y, depth_scale=10):
depth_intrin = depth_frame.profile.as_video_stream_profile().intrinsics
depth_pixel = [pixel_x, pixel_y]
depth_image_point = rs.rs2_deproject_pixel_to_point(depth_intrin, depth_pixel, depth_frame.get_distance(pixel_x, pixel_y))
depth_relative_point = np.array([depth_image_point[0]*depth_scale, depth_image_point[1]*depth_scale, depth_image_point[2]*depth_scale])
return depth_relative_point

rospy.init_node('sub')
rospy.Subscriber('/camera/depth/image_rect_raw',Image, Image_callback)

rospy.spin()

pipeline.stop()
cv2.destroyAllWindows()
`

Whenever I run this code, No device error occurred.

Traceback (most recent call last): File "/home/park/catkin_ws/src/robot/src/sub.py", line 27, in <module> pipeline.start(config) RuntimeError: No device connected

Realsense Camera D405.

[MartyG-RealSense]

Hi @kevinDrawn Hi @kevinDrawn I note that you are using a D405 with the ROS1 wrapper (as you are using rs_camera.launch). The ROS2 wrapper ros2-development supports the D405 camera model but the ROS1 wrapper does not by default. A RealSense ROS user published a custom "forked" version of the ROS1 wrapper at https://github.com/rjwb1/realsense-ros that contains D405 support.

If you want to create your own D405-supporting fork version of the wrapper instead of using the pre-made one at the above link, the edits to make to the file include/realsense2_camera/constants.h are described at #2582 (comment) and #2297 (comment)

---

The RealSense ROS1 wrapper has an example ROS pyrealsense2 node script called show_center_depth.py that uses the /camera/depth/image_rect_raw to print the real-world distance of the coordinate at the center of the image.

https://github.com/IntelRealSense/realsense-ros/blob/ros1-legacy/realsense2_camera/scripts/show_center_depth.py

It uses intrinsics from the /camera/depth/camera_info topic to calculate the 3D position of a pixel whose range is obtained from the depth image, received from the /camera/depth/image_rect_raw topic.

[kevinDrawn]

Thank you a lot!

[MartyG-RealSense]

You are very welcome! :)

[MartyG-RealSense]

Hi @kevinDrawn Do you require further assistance with this case, please? Thanks!

[MartyG-RealSense]

Case closed due to no further comments received.