Releases: robmarkcole/HASS-Deepstack-object
Bugfix saving images
Fixes an issue where images were saved regardless of target detection. Also adds a processed image with a constant filename to make it easy to display on the front end
Adds saving image with bounding box for object detection
Adds event image_processing.file_saved
and saving of images with bounding boxes if save_file_folder
is configured. An example automation is provided in the readme to send a notification with the saved image.
Adds image_processing.object_detected event
This release adds an event image_processing.object_detected
which is fired for each object detected.
The event payload includes:
classifier
: the classifier (i.e. deepstack_object)entity_id
: the entity id responsible for the eventobject
: the object detectedconfidence
: the confidence in detection in the range 0 - 1 where 1 is 100% confidence.
An example automation is provided in the instructions showing how a notification can be sent when a person
is detected
Updates for HA 0.92
Adds init and manifest files
Change file structure
Change file structure to adhere to new requirements for layout of custom components
v0.4
Adds service to teach (or register) faces with deepstack face
v0.3
Face recognition replaces gender detection
v0.2
Adds object detection
v0.1
Initial with basic functionality