[FEAT] min_area per camera which is responsible for ignoring detected areas due to its small size

[nordeep]

Is your feature request related to a problem? Please describe. It would be great to have the possibility to specify min_area: per camera as in main frigate: section. For now we have min_area: per camera for detect: but it's not the same. min_area: per camera for detect: is check to consider to match: or unknown:.

Describe the solution you'd like For now I have few cameras with high min_area: because those cameras are in front and at the level of face I don't want to overload detector with small area. But some wide cameras don't reach min_area:

cameras:
  front-door:
    min_area: 10000

[NickM-27]

You can already do this:

cameras:

  backyard_cam:
    detect:
      match:
        save: true
        confidence: 85
        min_area: 3000
      unknown:
        min_area: 2000
        confidence: 75
        
  front_doorbell_cam:
    detect:
      match:
        save: true
        confidence: 90
        min_area: 6500
      unknown:
        min_area: 4000
        confidence: 75
        purge: 48

[nordeep]

Sorry, but this is not the same as set min_area: here:

# frigate settings (default: shown below)
frigate:
  url:

  # if double take should send matches back to frigate as a sub label
  # NOTE: requires frigate 0.11.0+
  update_sub_labels: false

  # stop the processing loop if a match is found
  # if set to false all image attempts will be processed before determining the best match
  stop_on_match: true

  # ignore detected areas so small that face recognition would be difficult
  # quadrupling the min_area of the detector is a good start
  # does not apply to MQTT events
  min_area: 0

In frigate: section we specify - whether the detector will process image. In detect: section we specify - whether the process of detection will match or not.