viseron

Breaking changes

• Detectors now supply their own configuration validators which might break existing configs.
  If you used type: edgetpu and supplied an unsupported configuration option, such as suppression,
  Viseron would think that was okay. That is no longer the case and has to be removed.
  The README is updated accordingly.
• A new config block has been created for Home Assistant Discovery. It is still enabled by default but if you previously set your own discovery_prefix you now have to move this under home_assistant like this:

  mqtt:
    broker: <ip address or hostname of broker>
    port: <port the broker listens on>
    home_assistant:
      discovery_prefix: <yourcustomprefix>
  

• Pretty much all MQTT topics have changed. Have a look at the updated documentation to see how they are structured.

Changes and new Features

• Face detection is here! See the documentation for how to set this up.
  Here is an example config to get you started

  object_detection:
    labels:
      - label: person
        confidence: 0.8
        post_processor: face_recognition
  
  post_processors:
    face_recognition:
  

• Unique ID and Device registry information is now included in the Home Assistant discovery setup.
  This means that you can now customize the entities created in Home Assistant via its interface.
• The binary sensors created for each label now includes a count attribute which tells you the number of detected objects.
• Include support for h265 in ffmpeg for all containers
• vaapi and generic image now default to YOLOv3 model.
  Previously the default was YOLOv3-tiny, which is a lot faster but very inaccurate.
  If you wanna go back to using the tiny version for the sake of reduced CPU you can change these settings:

  object_detection:
    model_path: /detectors/models/darknet/yolov3-tiny.weights
    model_config: /detectors/models/darknet/yolov3-tiny.cfg
  

• Labels drawn on the image that is sent when publish_image: true is now clearer
• Adds support for PCIe based EdgeTPUs
• Adds a new configuration option called log_all_objects which is placed under object_detection

  object_detection:
    log_all_objects: true
  

  If set to true, all found objects will be logged if loglevel is DEBUG (this is the behaviour today). The default is false which will only log objects that pass labels filters.
• Uses ffprobe to get stream information instead of OpenCV.
• The automatic codec picker is now smarter and will try to find a matching hwaccel decoder codec for the source.
  Right now it only knows about h264 and h265 but this can be expanded in the future.
• New config option ffmpeg_loglevel under cameras. Use this to debug your camera decoding command.
• New config option ffmpeg_recoverable_errors under cameras. See the README for a detailed explanation.

Fixes

• Fixes an error when face_recognition folder is improperly structured
• EdgeTPU not being properly installed in the RPi image
• EdgeTPU not being loaded properly
• Fix a crash when publish_image: true and trigger_detector: false
• Fix interval under motion_detection and object_detection not allowing floats

Docker images are available on Docker Hub
roflcoopter/viseron:1.6.0
roflcoopter/viseron-cuda:1.6.0
roflcoopter/viseron-vaapi:1.6.0
roflcoopter/viseron-rpi:1.6.0

Changes and new Features

• Adds support for PCIe based EdgeTPUs
• Adds a new configuration option called log_all_objects which is placed under object_detection

  object_detection:
    log_all_objects: true
  

  If set to true, all found objects will be logged if loglevel is DEBUG (this is the behaviour today). The default is false which will only log objects that pass labels filters.
• Uses ffprobe to get stream information instead of OpenCV.
• The automatic codec picker is now smarter and will try to find a matching hwaccel decoder codec for the source.
  Right now it only knows about h264 and h265 but this can be expanded in the future.

Fixes

• Fixes an error when face_recognition folder is improperly structured
• EdgeTPU not being properly installed in the RPi image
• EdgeTPU not being loaded properly

Docker images are available on Docker Hub
roflcoopter/viseron:1.6.0b2
roflcoopter/viseron-cuda:1.6.0b2
roflcoopter/viseron-vaapi:1.6.0b2
roflcoopter/viseron-rpi:1.6.0b2

Breaking changes

• Detectors now supply their own configuration validators which might break existing configs.
  If you used type: edgetpu and supplied an unsupported configuration option, such as suppression,
  Viseron would think that was okay. That is no longer the case and has to be removed.
  The README is updated accordingly.
• A new config block has been created for Home Assistant Discovery. It is still enabled by default but if you previously set your own discovery_prefix you now have to move this under home_assistant like this:

  mqtt:
    broker: <ip address or hostname of broker>
    port: <port the broker listens on>
    home_assistant:
      discovery_prefix: <yourcustomprefix>
  

• Pretty much all MQTT topics have changed. Have a look at the updated documentation to see how they are structured.

Changes and new Features

• Face detection is here! See the documentation for how to set this up.
  Here is an example config to get you started

  object_detection:
    labels:
      - label: person
        confidence: 0.8
        post_processor: face_recognition
  
  post_processors:
    face_recognition:
  

• Unique ID and Device registry information is now included in the Home Assistant discovery setup.
  This means that you can now customize the entities created in Home Assistant via its interface.
• The binary sensors created for each label now includes a count attribute which tells you the number of detected objects.
• Include support for h265 in ffmpeg for all containers
• vaapi and generic image now default to YOLOv3 model.
  Previously the default was YOLOv3-tiny, which is a lot faster but very inaccurate.
  If you wanna go back to using the tiny version for the sake of reduced CPU you can change these settings:

  object_detection:
    model_path: /detectors/models/darknet/yolov3-tiny.weights
    model_config: /detectors/models/darknet/yolov3-tiny.cfg
  

• Labels drawn on the image that is sent when publish_image: true is now clearer

Fixes

• Fix a crash when publish_image: true and trigger_detector: false
• Fix interval under motion_detection and object_detection not allowing floats

Docker images are available on Docker Hub
roflcoopter/viseron:1.6.0b1
roflcoopter/viseron-cuda:1.6.0b1
roflcoopter/viseron-vaapi:1.6.0b1
roflcoopter/viseron-rpi:1.6.0b1

Lots of goodies in this one!

Breaking changes

• area for motion_detection is now a percentaged based value and needs to be changed from an int to a float.
  This means that if you change your width and/or height, area wont be affected.

Changes and new Features

• Masks can now be configured under each cameras motion_detection block.
  Masks are used to limit motion detection from running in a specified area.
  The configuration is similar to how zones are configured, here is an example:

  cameras:
    - name: name
      host: ip
      port: port
      path: /Streaming/Channels/101/
      motion_detection:
        area: 0.07
        mask:
          - points:
              - x: 0
                y: 0
              - x: 250
                y: 0
              - x: 250
                y: 250
              - x: 0
                y: 250
          - points:
              - x: 500
                y: 500
              - x: 1000
                y: 500
              - x: 1000
                y: 750
              - x: 300
                y: 750
  

  The masks are drawn on the image published over MQTT. they have an orange border with a black background with 70% opacity.

• A switch entity is now created in Home Assistant which can be used to arm/dsiarm a camera.
  When disarmed, the decoder stops completly so no system load will be used.

• Motion contours are now drawn on the image that is being published over MQTT.
  Dark purple contours are smaller than the configured area, while bright pink contours are larger than the configured area.

• New config option max_timeout under motion_detection which specifies the max number of seconds that motion alone is allowed to keep the recorder going.
  This is used to prevent never-ending recordings when motion detection is too sensitive

• New config option rtsp_transport under cameras. Change this if your camera doesnt support tcp.

• VA-API is now installed in the CUDA image for use with ffmpeg decoding/encoding

Fixes

• The binary sensor for motion detection is now properly set to on/off.

Docker images are available on Docker Hub
roflcoopter/viseron:latest
roflcoopter/viseron-cuda:latest
roflcoopter/viseron-vaapi:latest
roflcoopter/viseron-rpi:latest

Changes and new Features

• Zones are here! You can now configure zones for each camera and specify labels to track per zone.
  Here is an example:

  cameras:
    - name: name
      host: ip
      port: port
      path: /Streaming/Channels/101/
      zones:
        - name: zone1
          points:
            - x: 0
              y: 500
            - x: 1920
              y: 500
            - x: 1920
              y: 1080
            - x: 0
              y: 1080
          labels:
            - label: person
              confidence: 0.9
        - name: zone2
          points:
            - x: 0
              y: 0
            - x: 500
              y: 0
            - x: 500
              y: 500
            - x: 0
              y: 500
          labels:
            - label: cat
              confidence: 0.5
  

  A polygon will be drawn on the image using each point. Atleast 3 points have to be supplied.
  If you are using Home Assistant, Viseron will publish an image to the camera entity over MQTT
  with zones and objects drawn upon it.
  The drawing and publishing takes some processing power so it should only be used for debugging and tuning.

• A boatload of new binary sensors are now created, tracking objects and zones. Checkout the README for a detailed explanation.

• Allows a logging block to be entered per camera.

• Logging from motion detection is now named per camera.

• The logger for the recorder is now named per camera.

• You can now set log level individually for motion_detection and object_detection, either globally or for each camera.

• New config option publish_image.
  If enabled, Viseron will publish an image to the MQTT camera entity with objects and zones drawn upon it.

• You can now specify width, height and fps individually in the camera config.

• New config option triggers_recording for labels, if set to false, only the binary sensors in mqtt will update but no recording will start. This works on all labels configs (global object detector, per camera or per zone)

• Recorded videos will now be saved under a specific folder per camera.

Fixes

• TFLite is now properly installed in the generic, vaapi and cuda image for EdgeTPU support.
• Fixed motion detection not respecting timeout.

Docker images are available on Docker Hub
roflcoopter/viseron
roflcoopter/viseron-cuda
roflcoopter/viseron-vaapi
roflcoopter/viseron-rpi

Changes and new Features

• Allows a logging block to be entered per camera.
• New config option publish_image.
  If enabled, Viseron will publish an image to the MQTT camera entity with objects and zones drawn upon it. If a tracked object is in a zone the zone will turn green.
• You can now specify width, height and fps individually in the camera config.

Fixes

• The binary sensors created for zones and labels in zones are now properly turned on/off.

Docker images are available on Docker Hub
roflcoopter/viseron:dev
roflcoopter/viseron-cuda:dev
roflcoopter/viseron-vaapi:dev
roflcoopter/viseron-rpi:dev

Changes and new Features

• Zones are here! Functionality is somewhat limited atm but i need some testers on this as its quite a big refactor.
  You can now configure zones for each camera and specify labels to track per zone.
  This is not reflected in the documentation just yet, but here is an example:

  cameras:
    - name: name
      host: ip
      port: port
      path: /Streaming/Channels/101/
      zones:
        - name: zone1
          points:
            - x: 0
              y: 500
            - x: 1920
              y: 500
            - x: 1920
              y: 1080
            - x: 0
              y: 1080
          labels:
            - label: person
              confidence: 0.9
        - name: zone2
          points:
            - x: 0
              y: 0
            - x: 500
              y: 0
            - x: 500
              y: 500
            - x: 0
              y: 500
          labels:
            - label: cat
              confidence: 0.5
  

  A polygon will be drawn on the image using each point. Atleast 3 points have to be supplied.
  If you are using Home Assistant Viseron will publish an image to the camera entity over MQTT
  with zones and objects drawn upon it.
  The drawing and publishing takes some processing power so in the coming beta releases this will be configurable.

  A few new binary sensors will also be created, one for each zone and one for each label in the zone.
  The zone binary sensor will turn on when atleast one tracked object is in the zone.
  The label binary sensor will turn on when atleast one matching object is in the zone.

Docker images are available on Docker Hub
roflcoopter/viseron:dev
roflcoopter/viseron-cuda:dev
roflcoopter/viseron-vaapi:dev
roflcoopter/viseron-rpi:dev

Changes and new Features

• Support for mjpeg streams. To use it, add stream_format: mjpeg to your camera configuration.
  This is not fully tested since i dont have access to an mjpeg camera.

Fixes

• Fixed interval for object detector and motion detector. It now allows floats.
• No longer uses infinite memory when lookback: 0.
• Fixed user supplied codec being overwritten bu default value.

Docker images are available on Docker Hub
roflcoopter/viseron
roflcoopter/viseron-cuda
roflcoopter/viseron-vaapi
roflcoopter/viseron-rpi

Changes and new Features

• You can now make use of a secrets.yaml to substitute values in your config file.
• Added some benchmarks to the README

Fixes

• Fixes issue with default motion detection config causing decoder to crash

Docker images are available on Docker Hub
roflcoopter/viseron
roflcoopter/viseron-cuda
roflcoopter/viseron-vaapi
roflcoopter/viseron-rpi

Breaking changes
Object detecion config has changed significantly.
You now specify confidence and min/max height/width per label, example:

      labels:
        - label: person
          confidence: 0.9
        - label: truck

Changes

• You can now specify object_detection and motion_detection config per camera. See README for further explanation
• Refactored Darknet. It is now faster and easier to understand
• Logging has been improved significantly to make it easier to spot errors in the ffmpeg decoder command

Fixes

• Fixes an error specifying codec in camera config

Docker images are available on Docker Hub
roflcoopter/viseron
roflcoopter/viseron-cuda
roflcoopter/viseron-vaapi
roflcoopter/viseron-rpi

First release!
This is the first release of Viseron.
More features will be added as i go along.

Docker images are available on Docker Hub
roflcoopter/viseron
roflcoopter/viseron-cuda
roflcoopter/viseron-vaapi
roflcoopter/viseron-rpi

I hope you find this useful.