Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Object detected but clip does not include detected object (0.9.0-7B063A1)

See original GitHub issue

Describe the bug Event list show a detected object but when watching the recorded clip the object is not there.

Version of frigate 0.9.0-7b063a1

Config file Include your full config file wrapped in triple back ticks.

cameras:
#############################
#        PARKERING          #
#############################
  frigate_parkering:
    ffmpeg:
      inputs:
        - path: rtmp://192.168.1.216/bcs/channel0_main.bcs?channel=0&stream=1&user=admin&password=xxxxxxxx
          roles:
            - record
            - detect

    detect:
      width: 2560
      height: 1920
      fps: 5
      enabled: True
      max_disappeared: 25 

    objects:
      track:
        - person
        - djur
        - motorfordon
        - cykel
      filters:
        person:
          min_score: 0.7
        motorfordon:
          min_score: 0.6

    mqtt:
      enabled: True
      timestamp: False
      bounding_box: False
      crop: True
      height: 960
      quality: 100

    motion:
      threshold: 20
      mask:
        - 0,1920,347,1920,744,920,156,947,852,142,2560,129,2560,0,0,0
    
   zones:
      street_area:
        # HI-RES
        coordinates: 1093,386,1195,445,2464,496,2428,89,781,48,586,478,771,403
        objects:
          - person
          - djur
          - cykel
          - motorfordon
      parking_area:
        coordinates: 2560,1920,331,1920,748,916,147,932,556,451,579,567,400,898,923,880,1177,487,2560,543      
        objects: 
          - motorfordon
          - person
          - cykel
          - djur
        filters:
          person:
            threshold: 0.7
          motorfordon:
            threshold: 0.7
####################################
#                                  #
#         MAIN SETTINGS            #
#                                  #
####################################

detectors:
  coral1:
    type: edgetpu
    device: 'usb:0'
  coral2:
    type: edgetpu
    device: 'usb:1'

mqtt:
  host: 192.168.1.121
  topic_prefix: frigate
  user: home-assistant-server
  password: xxxxx
  stats_interval: 20

ffmpeg:
  hwaccel_args:
    - -hwaccel
    - qsv
    - -qsv_device
    - /dev/dri/renderD128

  output_args:
    record:
      - -f
      - segment
      - -segment_time
      - 10
      - -segment_format
      - mp4
      - -reset_timestamps
      - 1
      - -strftime
      - 1
      - -c:v
      - copy

  input_args:
    - '-avoid_negative_ts'
    - make_zero
    - '-fflags'
    - nobuffer
    - '-flags'
    - low_delay
    - '-strict'
    - experimental
    - '-fflags'
    - +genpts+discardcorrupt
    - '-use_wallclock_as_timestamps'
    - '1' 
  
detect:
  max_disappeared: 20

motion:
  threshold: 20
  contour_area: 100
  delta_alpha: 0.2
  frame_alpha: 0.2
  frame_height: 180

snapshots:
  bounding_box: true
  clean_copy: false
  crop: true
  enabled: true
  height: 600
  quality: 100
  retain:
    default: 6
  timestamp: false

objects:
  track:
    - djur
    - person
    - cykel
    - motorfordon
    - råtta
  filters:
    djur:
      threshold: 0.7
    person:
      threshold: 0.7
    cykel:
      threshold: 0.7
    motorfordon:
      threshold: 0.7
    råtta:
      threshold: 0.65

rtmp: 
  enabled: False

record:
  enabled: True
  retain_days: 0
  events:
    max_seconds: 300
    pre_capture: 5
    post_capture: 5
    retain:
      default: 6

birdseye:
  enabled: true
  height: 600
  width: 800
  mode: motion
  quality: 10


logger:
  default: info

Frigate container logs

Include relevant log output here

Frigate stats

{
	"detection_fps": 0.0,
	"detectors": {
		"coral1": {
			"detection_start": 0.0,
			"inference_speed": 11.94,
			"pid": 218
		},
		"coral2": {
			"detection_start": 0.0,
			"inference_speed": 11.65,
			"pid": 220
		}
	},
	"frigate_baksida": {
		"camera_fps": 5.1,
		"capture_pid": 267,
		"detection_fps": 0.0,
		"pid": 243,
		"process_fps": 5.1,
		"skipped_fps": 0.0
	},
	"frigate_entren": {
		"camera_fps": 4.1,
		"capture_pid": 249,
		"detection_fps": 0.0,
		"pid": 234,
		"process_fps": 4.1,
		"skipped_fps": 0.0
	},
	"frigate_inne_garage": {
		"camera_fps": 4.1,
		"capture_pid": 274,
		"detection_fps": 0.0,
		"pid": 246,
		"process_fps": 4.1,
		"skipped_fps": 0.0
	},
	"frigate_lillstugan": {
		"camera_fps": 4.1,
		"capture_pid": 257,
		"detection_fps": 0.0,
		"pid": 240,
		"process_fps": 4.1,
		"skipped_fps": 0.0
	},
	"frigate_parkering": {
		"camera_fps": 5.0,
		"capture_pid": 248,
		"detection_fps": 0.0,
		"pid": 229,
		"process_fps": 5.0,
		"skipped_fps": 0.0
	},
	"frigate_pathway": {
		"camera_fps": 4.1,
		"capture_pid": 260,
		"detection_fps": 0.0,
		"pid": 241,
		"process_fps": 4.1,
		"skipped_fps": 0.0
	},
	"frigate_pryttelverkstad": {
		"camera_fps": 4.0,
		"capture_pid": 270,
		"detection_fps": 0.0,
		"pid": 244,
		"process_fps": 4.0,
		"skipped_fps": 0.0
	},
	"frigate_tomten": {
		"camera_fps": 4.1,
		"capture_pid": 254,
		"detection_fps": 0.0,
		"pid": 238,
		"process_fps": 4.1,
		"skipped_fps": 0.0
	},
	"service": {
		"storage": {
			"/dev/shm": {
				"free": 6400.9,
				"mount_type": "tmpfs",
				"total": 6442.5,
				"used": 41.5
			},
			"/media/frigate/clips": {
				"free": 117205.5,
				"mount_type": "ext3",
				"total": 125488.4,
				"used": 1881.1
			},
			"/media/frigate/recordings": {
				"free": 117205.5,
				"mount_type": "ext3",
				"total": 125488.4,
				"used": 1881.1
			},
			"/tmp/cache": {
				"free": 22861.7,
				"mount_type": "overlay",
				"total": 58275.8,
				"used": 32423.3
			}
		},
		"uptime": 6824,
		"version": "0.9.0-7b063a1"
	}
}

FFprobe from your camera

Run the following command and paste output below

ffprobe <stream_url>

Screenshots Missing object_2

Computer Hardware

OS: [e.g. Ubuntu, Windows]
Install method: [e.g. Addon, Docker Compose, Docker Command]
Virtualization: [e.g. Proxmox, Virtualbox]
Coral Version: [e.g. USB, PCIe, None]
Network Setup: [e.g. Wired, WiFi]

Camera Info:

Manufacturer: Reolink
Model: 510A
Resolution: 2560x1920
FPS: 5

Additional context Add any other context about the problem here.

Issue Analytics

State:
Created 2 years ago
Comments:34 (15 by maintainers)

Top GitHub Comments

2reactions

blakeblackshearcommented, Sep 16, 2021

I see the same thing and I should be able to reproduce it and get a fix out in a few days.

1reaction

blakeblackshearcommented, Sep 22, 2021

Can you look at the timestamps on the recording segments and see if there are some missing based on the time range listed for the event? I’m trying to figure out if the recordings of the event were removed incorrectly or if the recordings are just out of sync with the event times.

Top Results From Across the Web

Option to create clips on motion where objects are not ...

I can use Zoneminder's motion detection features and alter the appropriate detection thresholds to record clips in addition to having Frigate ...

Configure a Video Analytics rule - Knowledge Base - Alarm.com

Note: Clips included in the Other category are when the camera detects an object but is unable to determine which type of object...

Setting Up Classified Object Motion Detection

Motion events can be used when searching recorded video, or to trigger notifications and rules. In the New Task menu , click Site...

Volumetric Features for Video Event Detection

Unfortunately, these conditions do not hold in typical real-world videos due to the presence of multiple moving objects and scene clutter. Similarly, the ......

Masks

Keep in mind that this mask only prevents motion from being detected and does not prevent objects from being detected if object detection...