[Detector Support]: Detect when Fence/Gate/Door will open/Close

[berndthiel]

Describe the problem you are having

Hi Team, i have a challenge in an automation i wanna create and hope you can help me to fix this :)

when

1. 0 cars in carport
2. when fance/gate is open
3. send notification

for 2) i dont have a solution, cause i dont understand how to use mask, zone and detection to identify this,
can someone help me to code this?

so i wanna combine data from camera parking, camera carport and camera housedoor

in camera parking i see is fence/gate open
fyi, i use FAAC 741 which doesnt have a feature/api to see the status, so i have to fix this by pictures i have e.g. "pic1-open", "pic2-closed"

Version

newest

Frigate config file

# MQTT settings
mqtt:
  host: core-mosquitto
  port: 1883
  topic_prefix: frigate
  client_id: frigate
  user: mqtt-user
  password: mqtt-password
  stats_interval: 60

# ui
detectors:
  ov:
    type: openvino
    device: CPU

model:
  width: 300
  height: 300
  input_tensor: nhwc
  input_pixel_format: bgr
  path: /openvino-model/ssdlite_mobilenet_v2.xml
  labelmap_path: /openvino-model/coco_91cl_bkgr.txt

# Live view settings
live:
  height: 720
  quality: 10

# Birdseye
birdseye:
  enabled: false
  mode: continuous

objects:
  track:
    - person
    - car
    - cat
    - dog
    - horse
    - bicycle
  filters:
    person:
      threshold: 0.72

# Record, Detect, Snapshots
record:
  enabled: true
  retain:
    days: 0
    mode: all
  events:
    retain:
      default: 30 # <- number of days to keep event recordings
      mode: all

detect:
  width: 600
  height: 300
  fps: 5

snapshots:
  enabled: true
  timestamp: true
  bounding_box: true
  crop: true
  retain:
    default: 30

### go2rtc ###
go2rtc:
  streams:
    # JideTech
    parking:
      - rtsp://admin:thielcamPL112B@192.168.100.11/cam1/mpeg4
    #H.VIEW
    housedoor:
      - rtsp://admin:xxx@192.168.100.12:554/live/main
    #Reolink
    carport_main:
      - rtsp://admin:xxx@192.168.100.14:554/h264Preview_01_main#video=copy#audio=copy#audio=opus
    carport_sub:
      - rtsp://admin:xxx@192.168.100.14:554/h264Preview_01_sub#video=copy#audio=copy#audio=opus
    garden_main:
      - rtsp://admin:xxx@192.168.100.15:554/h264Preview_01_main#video=copy#audio=copy#audio=opus
    garden_sub:
      - rtsp://admin:xxx@192.168.100.15:554/h264Preview_01_sub#video=copy#audio=copy#audio=opus

##############################################################

cameras:
# parking
  parking:
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/parking
          input_args: preset-rtsp-restream-low-latency
          roles:
            - detect
            - record

    motion:
      threshold: 45
      contour_area: 25
      improve_contrast: 'true'
      mask: 
        -0.002,0.361,0.279,0.363,0.275,0.228,0.644,0.238,0.635,0.358,0.99,0.363,1,-0.002,0.002,-0.002
    zones:
      parking_einfahrt:
        coordinates: 0.281,0.369,0.663,0.374,0.803,0.444,0.166,0.475
        loitering_time: 0
      parking_parkplatz:
        coordinates: 0.162,0.475,0.105,0.581,0.152,0.992,0.994,0.989,0.989,0.464,0.82,0.441
        loitering_time: 0
    objects:
      mask: 
        0.005,0,0.003,0.363,0.278,0.363,0.274,0.228,0.644,0.241,0.635,0.363,0.75,0.421,0.844,0.441,0.993,0.461,0.993,0.008
  housedoor:
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/housedoor
          input_args: preset-rtsp-restream-low-latency
          roles:
            - detect
            - record

    motion:
      threshold: 41
      contour_area: 10
      improve_contrast: 'true'
      mask: 0.005,0.075,0.194,0.103,0.205,0.04,0.012,0.023
    zones:
      housedoor_001:
        coordinates: 0.47,0.65,0.505,0,0.165,0,0.17,0.145
        inertia: 3
        loitering_time: 0
        objects:
          - person
          - car
          - bicycle
      housedoor_002:
        coordinates: 0.106,1,0.681,1,0.887,0.439,0.737,0.361,0.471,0.609,0.251,0.302,0,0.615,0,0.785

# carport
        inertia: 3
        loitering_time: 0
        objects:
          - person
          - dog
          - cat
    objects:
      filters:
        car: {}
      mask:
        - 0.002,0.441,0.062,0.411,0.206,0.887,0.121,0.993,0.002,0.993
        - 0,0.098,0.21,0.123,0.213,0.015,0.006,0.003
  carport:
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/carport_main
          input_args: preset-rtsp-restream
          roles:
            - record
        - path: rtsp://127.0.0.1:8554/carport_sub
          input_args: preset-rtsp-restream
          roles:
            - detect

    motion:
      threshold: 50
      contour_area: 10
      improve_contrast: 'true'
      mask: 0.167,0.36,0.741,0.348,0.994,0.528,0.993,0.003,0,0,0,0.186,0.167,0.163
    zones:
      carport001:
        coordinates: 
          0.562,0.321,0.174,0.366,0.061,0.464,0,0.574,0,1,1,1,0.887,0.729,0.915,0.525,0.951,0.497,0.854,0.39,0.705,0.352
        loitering_time: 0
        objects:
          - car
          - person
          - dog
          - cat
        inertia: 3
    objects:
      filters:
        car: {}

# Garden
  garden:
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/garden_main
          input_args: preset-rtsp-restream
          roles:
            - record
        - path: rtsp://127.0.0.1:8554/garden_sub
          input_args: preset-rtsp-restream
          roles:
            - detect

    motion:
      threshold: 40
      contour_area: 39
      improve_contrast: 'true'
      mask: 0,0.931,0.201,0.547,0.561,0.492,0.591,0.568,1,0.693,1,0,0,0
    zones:
      garden_001:
        coordinates: 0.993,0.712,0.993,0.812,0.456,0.617,0.576,0.592
        inertia: 3
        loitering_time: 0
      garden_002:
        coordinates: 336,152,338,174,264,179,264,155
      garden_003:
        coordinates: 0.064,1,0.044,0.864,0.196,0.574,0.253,0.584
        inertia: 3
        loitering_time: 0
      garden_004:
        coordinates: 600,300,534,300,542,255,600,209
      garden003:
        coordinates: 0.26,0.58,0.225,0.654,0.437,0.608,0.426,0.562
        loitering_time: 0
        objects: person
    objects:
      filters:
        car:
          mask:
            - 0,0.887,0.199,0.539,0.198,0.414,0.134,0.441,0.003,0.514
            - 0.58,0.554,0.995,0.674,0.992,0.474,0.485,0.439
version: 0.14

docker-compose file or Docker CLI command

-

Relevant Frigate log output

-

Relevant go2rtc log output

-

Operating system

HassOS

Install method

HassOS Addon

Object Detector

Coral

Screenshots of the Frigate UI's System metrics pages

Any other information that may be helpful

No response

[NickM-27]

I don't think there is a "frigate" solution to this, outside software will be needed

[VACIndustries]

This would be great! I'm not sure it would be super difficult training a model for your personal environment. Doing it for the wider "everyone" would take a bit more training I think.

This is one of those up there for those leaving the Google environment. I wouldn't be surprised if other cloud systems do similar