How to start camera stream on boot with systemd service – NVIDIA Jetson Tutorial

This tutorial demonstrates how to automatically launch a Python-based camera streaming application on boot using a systemd service on NVIDIA Jetson devices.

Prerequisites

• NVIDIA Jetson board (e.g., Orin NX, Nano, Xavier)
• Your custom Python GStreamer/WebRTC camera application
• Access to the terminal and root privileges

Step 1: Prepare the Python Script

Ensure your Python streaming script is available on the Jetson device. Below is the exact script used in this tutorial camera_stream.py:

 Note: Make sure to set the correct IP address for the device that will be consuming the stream. In this case 10.42.0.1 is used.

#!/usr/bin/env python3

import gi
import signal
import sys

gi.require_version('Gst', '1.0')
gi.require_version('GLib', '2.0')
from gi.repository import Gst, GLib

Gst.init(None)
loop = GLib.MainLoop()

def on_message(bus, message, loop):
    msg_type = message.type

    if msg_type == Gst.MessageType.ERROR:
        err, debug = message.parse_error()
        print(f"Error: {err}, Debug: {debug}")
        loop.quit()

def signal_handler(sig, frame):
    print("Interrupt received, stopping pipeline...")
    loop.quit()

# Handle SIGINT (Ctrl+C) and SIGTERM (systemd stop)
signal.signal(signal.SIGINT, signal_handler)
signal.signal(signal.SIGTERM, signal_handler)

pipeline = Gst.parse_launch(
    "nvarguscamerasrc sensor-id=0 ! "
    "nvv4l2h264enc bitrate=4000000 ! "
    "h264parse ! rtph264pay config-interval=1 pt=96 ! "
    "udpsink host=10.42.0.1 port=5000"
)

bus = pipeline.get_bus()
bus.add_signal_watch()
bus.connect("message", on_message, loop)

pipeline.set_state(Gst.State.PLAYING)

try:
    print("Pipeline started. Streaming...")
    loop.run()
finally:
    print("Shutting down pipeline.")
    pipeline.set_state(Gst.State.NULL)
    bus.remove_signal_watch()

Save this as /usr/local/bin/camera_stream.py.

Make sure all required GStreamer plugins are installed and working. You can test this script manually before proceeding.

python3 /usr/local/bin/camera_stream.py

Step 2: Create a systemd Service File

Use the following systemd unit file to run the Python script at boot:

[Unit]
Description=Jetson camera UDP stream on boot
After=nvargus-daemon.service
Requires=nvargus-daemon.service

[Service]
ExecStart=/usr/bin/python3 /usr/local/bin/camera_stream.py
Restart=on-failure
StandardOutput=journal
StandardError=journal

Save this file as /etc/systemd/system/camera-stream.service.

Step 3: Enable and Start the Service

To enable and start the service on boot, run the following commands:

sudo systemctl daemon-reload
sudo systemctl enable camera-stream.service
sudo systemctl start camera-stream.service

To check the service status:

systemctl status camera-stream.service

To view logs:

journalctl -u camera-stream.service

Step 4: Testing auto camera stream feature

In order to test this feature, run the following GStreamer pipeline on the device that will preview the camera stream put by the NVIDIA Jetson device.

gst-launch-1.0 -v udpsrc port=5000 ! application/x-rtp, encoding-name=H264, payload=96 ! rtph264depay ! avdec_h264 ! videoconvert ! autovideosink

Then, reboot or power-cycle the board.

Finally, make sure that the stream is shown as expected on the client device.

Need Further Support?

📞 Book Consultation Call: Show Calendar!

📩 Contact Via Email: support@proventusnova.com

🌐 Visit Our Website: ProventusNova.com