[AI Router] Setting Up Frigate Cameras

This guide will walk you step by step through setting up cameras in Frigate, so you can stream IP cameras into Frigate and start using powerful AI object detection features.

 

 

 

1. Prerequisites

    Before setting up Frigate cameras, make sure you have completed the following:

  • Frigate is installed and you can log in. See the previous guide Installing Frigate for installation and first login.
  • Your IP cameras are connected to your router, RTSP protocol is enabled, and you have set up a dedicated RTSP username and password. Refer to your camera’s documentation to confirm RTSP support. The RTSP username and password are set in the camera’s settings and are separate from your IPCam cloud account.

 

2. Get Your Camera’s RTSP Stream URL

    Set up your network camera (IP Camera) to enable RTSP:

  • Make sure your camera is successfully connected to your local network (via Wi-Fi or Ethernet to your router), and it’s recommended to set a static IP address.
  • Log in to the camera’s management interface and enable RTSP (Real Time Streaming Protocol). This is required for Frigate to access the video stream.
  • Note the camera’s RTSP stream URL. This is the key for Frigate to connect to your camera. The format is usually rtsp://<username>:<password>@<camera_ip_address>/<stream_path>。

 

3. Edit the Frigate Configuration File

    Go to the Frigate management interface and click the Configuration editor in the left menu. Edit the configuration file as shown below. Replace Name the camera and RTSP URL in the cameras section with your camera’s name and RTSP stream URL. RTSP URL formats vary by brand, so check your camera’s documentation for the correct format. <username> and <password> are the dedicated RTSP account credentials you set in the camera, and <IPCam IP> is the camera’s IP address. Be sure to replace these with your actual camera info.

mqtt:

      enabled: false

 

detectors:

    routernpu:

        type: npu

model:

      path: /cpu_model.tflite

      width: 320

      height: 320

 

cameras:

    AMCREST_Example: # <- Name the camera

          enabled: true

          ffmpeg:

              hwaccel_args: -c:v h264_v4l2m2m    # enable hardware acceleration

              input_args: preset-rtsp-generic    # enable hardware acceleration

              inputs:

                 - path: rtsp://<username>:<password>@<IPCam IP>:554/cam/realmonitor?channel=1&subtype=0 # <- RTSP URL

                   roles:

                      - detect

           detect:

               enabled: true

               width: 640

               height: 360

version: 0.15-1

 

    When finished, click Save & Restart in the upper right to save and restart Frigate.

 

4. Verify Camera Stream

    After restarting, go back to the Frigate homepage. You should see the camera feed you just added.

    If the feed doesn’t display, check that the RTSP URL matches your camera’s brand format.

 

5. Add a Second or More Cameras

    To add more cameras, specify a unique name and RTSP URL for each camera in the cameras section. <username> and <password> are the dedicated RTSP account credentials you set in the camera, and <IPCam IP> is the camera’s IP address. Be sure to replace these with your actual camera info.

mqtt: 
     enabled: false

 

detectors: 
    routernpu: 
        type: npu

 

model: 
      path: /cpu_model.tflite 
      width: 320 
      height: 320

 

cameras: 
    AMCREST_Example: # <- Name the camera 
          enabled: true 
          ffmpeg: 
              hwaccel_args: -c:v h264_v4l2m2m    # enable hardware acceleration

              input_args: preset-rtsp-generic    # enable hardware acceleration

              inputs:  
                - path: rtsp://<username>:<password>@<IPCam IP>:554/cam/realmonitor?channel=1&subtype=0 # <- RTSP URL 
                   roles: 
                      - detect 
          detect: 
              enabled: true # <---- disable detection until you have a working camera feed 
              width: 640 
              height: 360

 

Tapo_Example: # <- Name the camera 
   enabled: true 
   ffmpeg: 
       hwaccel_args: -c:v h264_v4l2m2m    # enable hardware acceleration

       input_args: preset-rtsp-generic    # enable hardware acceleration

       inputs:  
        - path: rtsp://<username>:<password>@<IPCam IP>:554/stream1 # <- RTSP URL 
          roles: 
             - detect 
   detect: 
       enabled: true 
       width: 640 
       height: 360

 

version: 0.15-1

    Once set up, you’ll see multiple camera feeds on the Frigate homepage

 

6. Advanced Settings

    Next, see Integrate Frigate with Home Assistant for more integration options. 

    You can also refer to the Frigate official documentation for more advanced configuration: https://docs.frigate.video/guides/getting_started#configuring-frigate

 

7. Frequently Asked Questions

   (1) What’s the difference between the CPU usage shown in the lower left and lower right corners of the Frigate screen?

  • The CPU usage in the lower left corner shows the total CPU usage of the Frigate container (or the whole system), including all Frigate operations such as object detection scheduling, web UI rendering, and the sum of all camera FFMPEG processing.
  • The CPU usage in the lower right corner refers only to the CPU resources used by FFMPEG to process the stream for that specific camera. This value varies depending on the camera’s resolution and whether hardware acceleration is enabled.
  • You can find more detailed system resource usage by clicking the gear icon in the lower left corner of Frigate, then selecting Settings -> System metrics.
  • Typically, when using one IPCam for event detection at 640x360 resolution with hardware acceleration enabled, the FFMPEG CPU usage in the lower right drops from 70% to 20%, and the overall container CPU usage in the lower left drops from 40% to below 15%.

 

 

 

 

How to get the (Utility / Firmware)?

You can download the latest drivers, software, firmware and user manuals in the ASUS Download Center.

If you need more information about the ASUS Download Center, please refer this link.