Frigate Dog, Cat & Raccoon detection issues

My Frigate setup does a fantastic detecting for “car” and “person”. So, I tried implementing dog, cat, and yes raccoon detection for my backyard grass area. The idea was as soon as a family of raccoons were detected, digging the grass for warms, the sprinkler valves would turn, to deter the raccoon family!

Unfortunately, I am having a hard time getting my cameras and Frigate to detect any animal the size of a raccoon. I have tried many changes with success.

Here is my frigate.yaml file for one camera7, showing different filter methods. I would appreciate your help. Thank you

detectors:
  coral:
    type: edgetpu
    device: usb

mqtt:
  host: IP
  user: USER_NAME
  password: PASSWORD


  #--------------------------------
  # <------ Name the camera 7 BEGIN
  #--------------------------------
  camera7:
    enabled: true
    ffmpeg:
      #hwaccel_args:  # recommended By ChatGPT. Remove if you see no difference
      #  - -hwaccel   # recommended By ChatGPT. Remove if you see no difference
      #  - qsv        # recommended By ChatGPT. Remove if you see no difference
      output_args:
        record: preset-record-generic-audio-copy
      inputs:
        - path: rtsp://admin:PASSWORD@IP:554/cam/realmonitor?channel=7&subtype=1
          input_args: preset-rtsp-restream
          roles:
            - detect
        - path: rtsp://admin:PASSWORD@IP:554:554/cam/realmonitor?channel=7&subtype=0
          roles:
            - record

    detect:
      width: 1270 # <- optional, by default Frigate tries to automatically detect resolution
      height: 720 # <- optional, by default Frigate tries to automatically detect resolution
      fps: 5
      stationary:
        interval: 20
        threshold: 50

    # object track ????
    objects:
      track:
        - person
        - dog
        - cat
        - raccoon  # Add raccoon to the list

      filters: # suggested config CPT
        person:
          min_area: 6000
          max_area: 150000
          min_score: 0.6
          threshold: 0.6
        dog:
          min_area: 4000
          max_area: 120000
          min_score: 0.5
          threshold: 0.6
        cat:
          min_area: 2000
          max_area: 80000
          min_score: 0.5
          threshold: 0.6
        raccoon:
          min_area: 3000
          max_area: 100000
          min_score: 0.5
          threshold: 0.6

#      filters:   # original config
#        person:
#          min_area: 5000
#          min_score: 0.5
#          threshold: 0.6
#        dog:
#          min_area: 5000
#          min_score: 0.5
#          threshold: 0.6
#        cat:
#          min_area: 5000
#          min_score: 0.5
#          threshold: 0.6
#        raccoon:  # Add filter for raccoon
#          min_area: 5000
#          min_score: 0.5
#          threshold: 0.6

    # Optional: Record configuration
    record:
      enabled: true
      retain:
        days: 3
        mode: active_objects #options: all motion active_objects
      events:
        pre_capture: 20
        post_capture: 20
        retain:
          default: 3
          mode: active_objects #options: All motion active_objects
          # Optional: Objects to save recordings for. (default: all tracked objects)
        objects:
          - person
          - dog
          - cat
          - raccoon
        #  Optional: Per object retention days
        #  objects: 4
        #  person: 7

    # Optional: Configuration for the jpg snapshots written to the clips directory for each event
    snapshots:
      enabled: true
      timestamp: true
      bounding_box: true
      crop: false
      retain:
        default: 3

    zones:
      back2_zone1:
        coordinates: 0.005,0.993,0,0.219,0.228,0.14,0.202,0.212,0.661,0.189,0.994,0.551,0.994,0.994
        loitering_time: 0
        objects: person
        inertia: 3
      back2_zone2:
        coordinates: 
          0.108,0.115,0.195,0.087,0.305,0.068,0.483,0.069,0.5,0.052,0.556,0.06,0.659,0.178,0.22,0.196,0.249,0.125,0.082,0.171
        loitering_time: 0
        objects:
          - person
          - dog
          - cat
          - raccoon
        inertia: 3
    motion:
      mask: 0.752,0.035,0.75,0.077,0.906,0.082,0.904,0.034
    review:
      alerts:
        required_zones:
          - back2_zone1
          - back2_zone2
      detections:
        required_zones:
          - back2_zone1
          - back2_zone2

raccoon is not a supported object on the default model. For dog and cat I’d suggest removing the min / max area filters if you are having trouble detecting them