[Support]: "Filters is very slow "

[Oli1549]

Describe the problem you are having

Filters are very slow, often 500ms+

Cant figure out why, as it didn't seem to do it when I first setup frigate.

im using a usb google coral TPU

Please see: https://i.imgur.com/j1epqVR.png

Version

0.14.1-f4f3cfa

What browser(s) are you using?

No response

Frigate config file

mqtt:
  enabled: true
  host: 192.168.1.3
  user: *
  password: *

record:
  enabled: true
  retain:
    days: 3
    mode: all
  events:
    retain:
      default: 30
      mode: motion

## logger:
 ## logs:
 ##   frigate.record: debug

detectors:
  coral:
    type: edgetpu
    device: usb


#Global Object Settings
  filters:
    person:
      min_area: 5000
      max_area: 100000
      threshold: 0.7

go2rtc:
  streams:
    Backyard:
      - rtsp://admin:*@192.168.1.80/Streaming/Channels/101/
    Backyard-sub:
      - rtsp://admin:*@192.168.1.80/Streaming/Channels/102/
    Front:
      - rtsp://admin:*@192.168.1.81/Streaming/Channels/101/
    Front-sub:
      - rtsp://admin:*@192.168.1.81/Streaming/Channels/102/
    Drive:
      - rtsp://admin:*@192.168.1.82/Streaming/Channels/101/
    Drive-sub:
      - rtsp://admin:*@192.168.1.82/Streaming/Channels/102/
    Lawn:
      - rtsp://admin:*@192.168.1.83/Streaming/Channels/101/
    Lawn-sub:
      - rtsp://admin:*@192.168.1.83/Streaming/Channels/102/




cameras:
  Backyard: # <------ Name the camera
    enabled: true
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/Backyard-sub # <----- The stream you want to use for detection
          input_args: preset-rtsp-restream
          roles:
            - detect
        - path: rtsp://127.0.0.1:8554/Backyard # <----- The stream you want to use for detection
          input_args: preset-rtsp-restream
          roles:
            - record
    detect:
      enabled: true # <---- disable detection until you have a working camera feed
      width: 1280
      height: 720
    objects:
      track:
        - person
        - bicycle
        - motorcycle
        - cat
        - dog
    snapshots:
      enabled: true

    motion:
      threshold: 60
      contour_area: 10
      improve_contrast: 'true'
  Front: # <------ Name the camera
    enabled: true
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/Front-sub # <----- The stream you want to use for detection
          input_args: preset-rtsp-restream
          roles:
            - detect
        - path: rtsp://127.0.0.1:8554/Front # <----- The stream you want to use for detection
          input_args: preset-rtsp-restream
          roles:
            - record
    detect:
      enabled: true # <---- disable detection until you have a working camera feed
      width: 1280
      height: 720
    objects:
      track:
        - person
        - bicycle
        - motorcycle
        - cat
        - dog

    snapshots:
      enabled: true

    motion:
      threshold: 70
      contour_area: 30
      improve_contrast: 'true'
  Drive: # <------ Name the camera
    enabled: true
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/Drive-sub # <----- The stream you want to use for detection
          input_args: preset-rtsp-restream
          roles:
            - detect
        - path: rtsp://127.0.0.1:8554/Drive # <----- The stream you want to use for detection
          input_args: preset-rtsp-restream
          roles:
            - record
    detect:
      enabled: true # <---- disable detection until you have a working camera feed
      width: 1280
      height: 720

    snapshots:
      enabled: true

    motion:
      threshold: 70
      contour_area: 10
      improve_contrast: 'true'
    objects:
      track:
        - person
        - car
  Lawn: # <------ Name the camera
    enabled: true
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/Lawn-sub # <----- The stream you want to use for detection
          input_args: preset-rtsp-restream
          roles:
            - detect
        - path: rtsp://127.0.0.1:8554/Lawn # <----- The stream you want to use for detection
          input_args: preset-rtsp-restream
          roles:
            - record
    detect:
      enabled: true # <---- disable detection until you have a working camera feed
      width: 1280
      height: 720
    objects:
      track:
        - person
        - bicycle
        - motorcycle
        - cat
        - dog
    snapshots:
      enabled: true

    motion:
      threshold: 30
      contour_area: 10
      improve_contrast: 'true'


version: 0.14

Relevant Frigate log output

-

Relevant go2rtc log output

-

FFprobe output from your camera

-

Frigate stats

https://imgur.com/a/nJvslgw

Install method

Docker Compose

docker-compose file or Docker CLI command

Ran through unraid docker.

--shm-size=256m --mount type=tmpfs,target=/tmp/cache,tmpfs-size=1000000000 --restart unless-stopped

Object Detector

Coral

Network connection

Wired

Camera make and model

Hikvision 4K x4

Screenshots of the Frigate UI's System metrics pages

No response

Any other information that may be helpful

No response

[hawkeye217]

You are missing objects: above filters: in your config. It should look like this:

#Global Object Settings
objects:
  filters:
    person:
      min_area: 5000
      max_area: 100000
      threshold: 0.7