Editing Introduction to a Self Managed Life: a 13 hour & 28 minute presentation by FUTO software (section)

==== 3.3 Create Frigate Configuration File ====

<ol style="list-style-type: decimal;">
<li>'''Create and Edit the <code>config.yml</code> File:'''
<ul>
<li><p>Create a <code>config/config.yml</code> file to define your cameras &amp; MQTT setup.</p></li>
<li><p>''I have provided a template below. Creating yml files is painful and very easy to mess up. So I provided a known-working file for you to start with.''</p></li>
<li><p>'''YOU WILL HAVE TO EDIT THE IP ADDRESSES, USERNAMES, AND PASSWORDS IN EACH PATH LINE TO THE URL OF YOUR ACTUAL CAMERA. YOUR CAMERAS WILL ALSO HAVE DIFFERENT URLS THAN MINE. I DID MOST OF THE WORK FOR YOU, BUT DON’T BE SO LAZY THAT YOU DON’T EVEN CHANGE THE CAMERA IPs &amp; USERNAMES &amp; PASSWORDS TO YOURS!'''</p></li>
<li><p>To find the RTSP URLs of your camera, you can install <code>nmap</code> on Ubuntu with:</p>
<pre>sudo apt install nmap -y</pre></li>
<li><p>Then you go to your terminal and type the following, replacing the IP address of <code>192.168.3.120</code> with the IP address of your camera:</p>
<pre>sudo nmap --script rtsp-url-brute -p 554 192.168.5.19
sudo nmap --script rtsp-url-brute -p 8554 192.168.5.19</pre></li>
<li><p>You will receive a list of stream URLs. Let’s say one of them is <code>&quot;rtsp://192.168.5.19/Streaming/Channels/101&quot;</code>.</p></li></ul>
</li></ol>

* You need to add your username &amp; password here. So <code>rtsp://192.168.5.19/Streaming/Channels/101</code> will become <code>rtsp://username:password@192.168.5.19/Streaming/Channels/101</code>.
* Test that this works in a video player like VLC. In VLC, go '''Media''' → '''Open Network Stream''' → '''Network URL''' → enter the URL → click '''Play'''.
* If it works, it can be entered into the <code>path</code> line and replace my URLs in the config file below.
* The first four lines are going to be for MQTT, which sends messages to Home Assistant so that Home Assistant can send alerts to your phone when someone tries to steal your catalytic converter.

<pre>mqtt:
  host: homeassistant.home.arpa  
  port: 1883
  user: louis
  password: passwordman

cameras:
  front_door_closeup:
    ffmpeg:
      inputs:
        - path: rtsp://CAMERAUSERNAMEGOESHERE:CAMERAPASSWORDGOESHERE@192.168.3.101:554/Streaming/Channels/101
          roles:
            - record
        - path: rtsp://CAMERAUSERNAMEGOESHERE:CAMERAPASSWORDGOESHERE@192.168.3.101:554/Streaming/Channels/102
          roles:
            - detect
      output_args:
        record: -f segment -segment_time 60 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c copy
    detect:
      width: 640
      height: 360
      fps: 20
    objects:
      track:
        - person
        - car
        - motorcycle
        - bird
        - cat
        - dog
        - horse
        - sheep
        - cow
        - bear
        - zebra
        - giraffe
        - elephant
        - mouse
      filters:
        person:
          mask: 570,299,545,0
        cat:
          min_score: 0.01
          threshold: 0.02
        dog:
          min_score: 0.01
          threshold: 0.02
        bird:
          min_score: 0.01
          threshold: 0.02
    motion:
      mask:
        - 473,0,21,156,53,317,140,312
    record:
      enabled: true
      events:
        pre_capture: 5
        post_capture: 5
        objects:
          - person
          - car
          - motorcycle
          - bird
          - cat
          - dog
          - horse
          - sheep
          - cow
          - bear
          - zebra
          - giraffe
          - elephant
          - mouse

  driveway:
    ffmpeg:
      inputs:
        - path: rtsp://CAMERAUSERNAMEGOESHERE:CAMERAPASSWORDGOESHERE@192.168.3.102:554/Streaming/Channels/101
          roles:
            - record
        - path: rtsp://CAMERAUSERNAMEGOESHERE:CAMERAPASSWORDGOESHERE@192.168.3.102:554/Streaming/Channels/102
          roles:
            - detect
      output_args:
        record: -f segment -segment_time 60 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c copy
    detect:
      width: 640
      height: 360
      fps: 20
    objects:
      track:
        - person
        - car
        - motorcycle
        - bird
        - cat
        - dog
        - horse
        - sheep
        - cow
        - bear
        - zebra
        - giraffe
        - elephant
        - mouse
      filters:
        car:
          min_score: 0.01
          threshold: 0.03
        cat:
          min_score: 0.01
          threshold: 0.02
        dog:
          min_score: 0.01
          threshold: 0.02
        bird:
          min_score: 0.01
          threshold: 0.02
    record:
      enabled: true
      events:
        pre_capture: 5
        post_capture: 5
        objects:
          - person
          - car
          - motorcycle
          - bird
          - cat
          - dog
          - horse
          - sheep
          - cow
          - bear
          - zebra
          - giraffe
          - elephant
          - mouse

  side_door_closeup:
    ffmpeg:
      inputs:
        - path: rtsp://CAMERAUSERNAMEGOESHERE:CAMERAPASSWORDGOESHERE@192.168.3.104:554/Streaming/Channels/101
          roles:
            - record
        - path: rtsp://CAMERAUSERNAMEGOESHERE:CAMERAPASSWORDGOESHERE@192.168.3.104:554/Streaming/Channels/102
          roles:
            - detect
      output_args:
        record: -f segment -segment_time 60 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c copy
    detect:
      width: 640
      height: 360
      fps: 20
    objects:
      track:
        - person
        - bird
        - cat
        - dog
        - horse
        - sheep
        - cow
        - bear
        - zebra
        - giraffe
        - elephant
        - mouse
      filters:
        car:
          min_score: 0.01
          threshold: 0.03
        cat:
          min_score: 0.01
          threshold: 0.02
        dog:
          min_score: 0.01
          threshold: 0.02
        bird:
          min_score: 0.70
          threshold: 0.75
    record:
      enabled: true
      events:
        pre_capture: 5
        post_capture: 5
        objects:
          - person
          - car
          - bird
          - cat
          - dog
          - horse
          - sheep
          - cow
          - bear
          - zebra
          - giraffe
          - elephant
          - mouse

  back_door_closeup:
    ffmpeg:
      inputs:
        - path: rtsp://CAMERAUSERNAMEGOESHERE:CAMERAPASSWORDGOESHERE@192.168.3.103:554/Streaming/Channels/101
          roles:
            - record
        - path: rtsp://CAMERAUSERNAMEGOESHERE:CAMERAPASSWORDGOESHERE@192.168.3.103:554/Streaming/Channels/102
          roles:
            - detect
      output_args:
        record: -f segment -segment_time 60 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c copy
    detect:
      width: 640
      height: 360
      fps: 20
    objects:
      track:
        - person
        - car
        - bird
        - cat
        - dog
        - horse
        - sheep
        - cow
        - bear
        - zebra
        - giraffe
        - elephant
        - mouse
      filters:
        car:
          min_score: 0.75
          threshold: 0.75
        cat:
          min_score: 0.01
          threshold: 0.02
        dog:
          min_score: 0.01
          threshold: 0.02
        bird:
          min_score: 0.01
          threshold: 0.02
    record:
      enabled: true
      events:
        pre_capture: 5
        post_capture: 5
        objects:
          - person
          - car
          - bird
          - cat
          - dog
          - horse
          - sheep
          - cow
          - bear
          - zebra
          - giraffe
          - elephant
          - mouse

  front_porch_wide_angle:
    ffmpeg:
      inputs:
        - path: rtsp://CAMERAUSERNAMEGOESHERE:CAMERAPASSWORDGOESHERE@192.168.3.106:554/Streaming/Channels/101
          roles:
            - record
        - path: rtsp://CAMERAUSERNAMEGOESHERE:CAMERAPASSWORDGOESHERE@192.168.3.106:554/Streaming/Channels/102
          roles:
            - detect
      output_args:
        record: -f segment -segment_time 60 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c copy
    detect:
      width: 640
      height: 360
      fps: 20
    objects:
      track:
        - person
        - car
        - motorcycle
        - bird
        - cat
        - dog
        - horse
        - sheep
        - cow
        - bear
        - zebra
        - giraffe
        - elephant
        - mouse
      filters:
        person:
          min_score: 0.8
          threshold: 0.8
        car:
          min_score: 0.6
          threshold: 0.7
        cat:
          min_score: 0.01
          threshold: 0.02
        dog:
          min_score: 0.01
          threshold: 0.02
        bird:
          min_score: 0.6
          threshold: 0.65
    record:
      enabled: true
      events:
        pre_capture: 5
        post_capture: 5
        objects:
          - person
          - car
          - motorcycle
          - bird
          - cat
          - dog
          - horse
          - sheep
          - cow
          - bear
          - zebra
          - giraffe
          - elephant
          - mouse

  fishcam:
    ffmpeg:
      inputs:
        - path: rtsp://louis:passwordroflcopter@192.168.3.120:554/stream1
          roles:
            - record
        - path: rtsp://louis:passwordroflcopter@192.168.3.120:554/stream1
          roles:
            - detect
      output_args:
        record: -f segment -segment_time 60 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c copy
    detect:
      width: 640
      height: 360
      fps: 20
    objects:
      track:
        - person
      filters:
        person:
          min_score: 0.3
          threshold: 0.3
    record:
      enabled: true
      events:
        pre_capture: 15
        post_capture: 15
        objects:
          - fish

database:
  path: /data/db/frigate.db
#version: 0.14</pre>
<blockquote>'''Note:''' For each camera, configure the RTSP inputs for recording and detection streams. Define output arguments, detection settings (e.g., width, height, fps), and tracked objects (e.g., person, car, bird, dog). You can set filters for specific objects, mask areas for motion detection, and enable event recording with pre-capture and post-capture times. Repeat for additional cameras as needed.
</blockquote>
<span id="running-frigate"></span>