profile = pipeline.start(config) , RuntimeError

[m0-hayashimanato]

Camera Model : L515
Operating System & Version : Linux (Ubuntu-22.04)
Language : Python (3.10.12 64-bit)

、、、
import pyrealsense2 as rs
import numpy as np
import cv2
import open3d

何かキーを押した時点のフレームの.pcdファイルを作成

realsenseL515をこのPCに接続すれば動作するはずです

if name=="main":

ストリーミングの設定

align = rs.align(rs.stream.color)
config = rs.config()
config.enable_stream(rs.stream.depth, 1024, 768, rs.format.z16, 30)
config.enable_stream(rs.stream.color, 1280, 720, rs.format.rgb8, 30)
pipeline = rs.pipeline()
profile = pipeline.start(config)
depth_sensor = profile.get_device().first_depth_sensor()
color_sensor = profile.get_device().first_color_sensor()

depth_sensor.set_option(rs.option.min_distance, 0.0500)

# get camera intrinsics
intr = profile.get_stream(rs.stream.color).as_video_stream_profile().get_intrinsics()
print(intr.width, intr.height, intr.fx, intr.fy, intr.ppx, intr.ppy)
pinhole_camera_intrinsic = open3d.camera.PinholeCameraIntrinsic(intr.width, intr.height, intr.fx, intr.fy, intr.ppx, intr.ppy)
print(pinhole_camera_intrinsic)

ストリーミング開始

while True:
    frames = pipeline.wait_for_frames()
    aligned_frames = align.process(frames)

    color_frame = aligned_frames.get_color_frame()
    color_image = np.array(color_frame.get_data())

    depth_frame = frames.get_depth_frame()
    depth_image = np.array(depth_frame.get_data())

    colored_depth_frame = rs.colorizer().colorize(depth_frame)
    colored_depth_image = np.array(colored_depth_frame.get_data())

    #画像中心の実距離(m)を表示
    colored_depth_image_dot = colored_depth_image
    colored_depth_image_dot[384-3:385+3, 512-3:513+3, :] = 255
    distance = depth_frame.get_distance(512, 384)
    cv2.putText(colored_depth_image_dot, str('%.5f' % distance), (512+20, 384), cv2.FONT_HERSHEY_COMPLEX, 1, (255, 255, 255), 2, cv2.LINE_AA)

    cv2.imshow('color', cv2.cvtColor(color_image, cv2.COLOR_RGB2BGR))
    cv2.imshow('depth', cv2.cvtColor(colored_depth_image_dot, cv2.COLOR_RGB2BGR))

    if cv2.waitKey(1) != -1:
        print('finish')
        break

何かのキーを押した瞬間のデプス画像、カラー画像、PCDを作成

depth_frame = aligned_frames.get_depth_frame()

f = np.array(depth_frame.get_data())
depth = open3d.geometry.Image(f)
color = open3d.geometry.Image(color_image)

depth_np = np.array(depth)
color_np = np.array(color)
color_np = cv2.cvtColor(color_np, cv2.COLOR_BGR2RGB)
#cv2.imwrite("./temp_color_np.png", color_np)
#cv2.imwrite("./temp_depth_np.png", depth_np)

rgbd = open3d.geometry.RGBDImage.create_from_color_and_depth(color, depth, convert_rgb_to_intensity = False)

pcd = open3d.geometry.PointCloud.create_from_rgbd_image(rgbd, pinhole_camera_intrinsic)
pcd.transform([[1,0,0,0],[0,-1,0,0],[0,0,-1,0],[0,0,0,1]])

open3d.io.write_point_cloud('pcd.pcd', pcd)
open3d.visualization.draw_geometries([pcd])

print("finish")

、、、

I am trying to execute the code shown above, but it will not run with the following error.

、、、
Traceback (most recent call last):
File "/home/bio/貂｡霎ｺ繧ｽ繝ｼ繧ｹ/make_data.py", line 17, in
profile = pipeline.start(config)
RuntimeError: No device connected
、、、

Realsense L515 is connected to a PC.
I would like to know the solution to this problem.

[MartyG-RealSense]

Hi @m0-hayashimanato Which version of the librealsense SDK do you have installed, please? Support for L515 was removed from the SDK in version 2.54.1 as it is a retired camera model. If you are using 2.54.1 or newer then I recommend installing version 2.51.1 so that the L515 can be detected.

[m0-hayashimanato]

Thank you, @MartyG-RealSense . I originally installed version 2.54.2, so I installed version 2.51.1 as advised.
However, when I run it in that state, I get the same error again. Is there something I need to do to change the version?

[MartyG-RealSense]

Have you installed the pyrealsense2 wrapper again for 2.51.1 please?

If you have, please try unplugging the micro-sized end of the USB cable from the base of the L515 camera, reversing the connector's orientation and plugging it back into the L515 (USB-C cables are two-way insertion at the micro-sized end). Does doing so make any difference?

[m0-hayashimanato]

How do I install the pyrealsense2 wrapper for 2.51.1？

At this time, there is no diffrernce by reversing the connector's orientation and plugging it back into the L515.

[MartyG-RealSense]

If you are using your L515 on a PC computer then you could install version 2.51.1 of pyrealsense2 from PyPi packages with the command below:

pip install pyrealsense2==2.51.1.4348

[m0-hayashimanato]

When I ran the command in the terminal, I got the following statement.

Defaulting to user installation because normal site-packages is not writeable ERROR: Could not find a version that satisfies the requirement pyrealsense2==2.51.1.4348 (from versions: 2.53.1.4623, 2.54.1.5216, 2.54.1.5217, 2.54.2.5684, 2.55.1.6486) ERROR: No matching distribution found for pyrealsense2==2.51.1.4348

[MartyG-RealSense]

It looks as though the 2.51 PyPi package might not exist any more.

In that case, installing librealsense 2.53.1 and then doing pip install pyrealsense2==2.53.1.4623 might be a solution. The 2.53.1 version of librealsense also has L515 compatibility.

The main reason I usually recommend 2.51.1 for use with L515 is because 2.53.1 is not suitable for use with the RealSense ROS wrapper, but if you are not using ROS with your RealSense camera then 2.53.1 would be worth trying.

[m0-hayashimanato]

pip show pyrealsense2 Name: pyrealsense2 Version: 2.53.1.4623 Summary: Python Wrapper for Intel Realsense SDK 2.0. Home-page: https://github.com/IntelRealSense/librealsense Author: Intel(R) RealSense(TM) Author-email: [email protected] License: Apache License, Version 2.0 Location: /home/bio/.local/lib/python3.10/site-packages Requires: Required-by:

I installed librealsense 2.53.1 and pyrealsense2==2.53.1.4623, but I got a similar error.
I even tried reversing the direction of the connectors, but to no avail.

[MartyG-RealSense]

Are you able to test whether the L515 is detectable in librealsense 2.53.1 by running the tool rs-enumerate-devices

[m0-hayashimanato]

Is rs-enumerate-devices command? When I ran it in the terminal, I got the following statement.
rs-enumerate-devices: command not found

[MartyG-RealSense]

If the tool and example programs were included in your librealsense installation then it should be in the usr/local/bin directory of Ubuntu.

[m0-hayashimanato]

It was not in that directory. What should I do?

[MartyG-RealSense]

If you use the command lsusb in the terminal then you should at least be able to find out if the camera is detectable by the computer.

[m0-hayashimanato]

When I ran lsusb, realsense was not detected.

[MartyG-RealSense]

lsusb is independent of the RealSense SDK and simply reads the details of devices directly from the USB port. So if lsusb cannot find the camera, it would suggest that the non-detection problem is not caused by your librealsense installation but by a problem with the camera hardware or the USB cable.

I note that reversing the orientation of your USB cable did not make a difference.

How long have you owned your L515 camera, please?

Has there been a period of time when you have stored the camera unused in a place such as a drawer or cupboard for weeks / months before you tried the camera and found it to be non-detectable?

[m0-hayashimanato]

I reversed the direction of the USB cable and it did not change.
I don't know the exact number of years we have owned the camera since it was handed down to me, but I think I have used it for a total of 6 years. There was also a period of time when the camera was stored without use for a few months.

[MartyG-RealSense]

In rare cases a RealSense camera can experience a hardware failure when stored inactive for months. The reason for it is unknown but it may be due to condensation forming inside the camera's casing due to the environmental conditions in the storage location.

There was once a case where such a camera started working again after a few weeks. It is suspected that it was in warm conditions during this period. In the majority of cases though, it is unfortunately an unfixable failure.

[m0-hayashimanato]

Thank you.
Is there a difference between librealsense and pyrealsense? I would like to know how to install librealsense because I may not have installed it.

[MartyG-RealSense]

librealsense is the "main" part of the RealSense software that enables the camera to be used. Pyrealsense2 is an optional add-on to librealsense that enables Python programs to control the camera.

There are two approaches to installing them.

1. Build librealsense and pyrealsense2 from source code at the same time.

2. Install librealsense first and then install pyrealsense2 separately afterwards.

---

The easiest approach is to build librealsense and pyrealsense2 at the same time using the 'libuvc_installation.sh' build script.

https://github.com/IntelRealSense/librealsense/blob/master/doc/libuvc_installation.md

I recommend editing line 46 of the script to use the following instruction:

cmake ../ -DFORCE_LIBUVC=true -DCMAKE_BUILD_TYPE=release -DBUILD_EXAMPLES=true -DBUILD_GRAPHICAL_EXAMPLES=true -DBUILD_PYTHON_BINDINGS:bool=true

https://github.com/IntelRealSense/librealsense/blob/master/scripts/libuvc_installation.sh#L46

Then run the edited script file. This will install librealsense with pyrealsense2 and the tools and examples included.

[m0-hayashimanato]

Thank you very much. I was able to install librealsense as you taught me.

[m0-hayashimanato]

The error is not gone yet, but what should I do next?

[m0-hayashimanato]

Sorry for the consecutive posts. I think the version of librealsense and pyrealsense I installed are different, is it ok?

[MartyG-RealSense]

Does it make a difference if you run your script with sudo admin permissions?

sudo python3 make_data.py

[m0-hayashimanato]

It was the same.

[MartyG-RealSense]

Is the camera still undetected if you use the test script below?

# First import the library
import pyrealsense2 as rs

# Create a context object. This object owns the handles to all connected realsense devices
pipeline = rs.pipeline()
pipeline.start()

try:
    while True:
        # Create a pipeline object. This object configures the streaming camera and owns it's handle
        frames = pipeline.wait_for_frames()
        depth = frames.get_depth_frame()
        if not depth: continue

        # Print a simple text-based representation of the image, by breaking it into 10x20 pixel regions and approximating the coverage of pixels within one meter
        coverage = [0]*64
        for y in range(480):
            for x in range(640):
                dist = depth.get_distance(x, y)
                if 0 < dist and dist < 1:
                    coverage[x//10] += 1

            if y%20 is 19:
                line = ""
                for c in coverage:
                    line += " .:nhBXWW"[c//25]
                coverage = [0]*64
                print(line)

finally:
    pipeline.stop()

[m0-hayashimanato]

Not detected

[MartyG-RealSense]

There is not much more that can be tried, unfortunately.

Here are a couple of further suggestions to test.

1. Start the depth and color streams in the RealSense Viewer first before launching the Python program (as you said that the camera was working fine in the Viewer). Is the camera still undetected by the Python program?

Or:

2. Change 'import pyrealsense2 as rs' to import pyrealsense2.pyrealsense2 as rs

[m0-hayashimanato]

Neither method 1 or 2 worked.

I was able to install librealsense using the method you taught me, and now the Realsense viewer opens when I use the command realsense-viewer on the terminal. However, it opens version 2.55.1. Is this a problem?

[MartyG-RealSense]

realsense-viewer 2.55.1 should not be able to show the L515 as support for L515 was removed in 2.54.1. So it is strange that your L515 is working with viewer 2.55.1.

[MartyG-RealSense]

If you installed librealsense with the libuvc_installation.sh script as I suggested then it may be my fault that you have 2.55.1, as this script will install the latest librealsense version.

https://github.com/IntelRealSense/librealsense/blob/master/scripts/libuvc_installation.sh

A workaround to install 2.53.1 may be to follow the procedure below.

1. Delete lines 37-39 from the libuvc_installation.sh script

https://github.com/IntelRealSense/librealsense/blob/master/scripts/libuvc_installation.sh#L37-L39

2. Download and extract the Source code zip file for 2.53.1 from the 'Assets' file list at the bottom of the page at the link below.

https://github.com/IntelRealSense/librealsense/releases/tag/v2.53.1

3. Navigate to the 'root' directory of the librealsense source code folder. This is the second one in: librealsense > librealsense, which has the list of folders and files in it.

4. Copy the edited libuvc_installation.sh script into this root folder and run the script

[m0-hayashimanato]

I'm sorry. I don't know what the root directory is.
When you say “This is the second one in: librealsense > librealsense, which has the list of folders and files in it.” do you mean the CMake folder?
Also, how do I run the script?

[MartyG-RealSense]

When the source code for 2.53.1 is extracted from the zip file, you will have a folder called librealsense-2.53.1. Inside that folder, there is a sub-folder also called librealsense-2.53.1. When this second folder is opened, the list of folders and files are there. This is the 'root' folder that the libuvc_installation.sh script file should be copied into.

So it is not the CMake folder.

The script file 'libuvc_installation.sh' should be able to be able to be launched in the Ubuntu terminal if you navigate to 'librealsense-2.53.1 > librealsense-2.53.1' within the librealsense source code folder, copy the libuvc_installation.sh file to that location and then input the commands below.

 chmod +x ./libuvc_installation.sh
./libuvc_installation.sh