Releases: robmarkcole/HASS-plate-recognizer
Internal update
Merge pull request #51 from robmarkcole/ISSUE50 Update image_processing.py
Adds Coordinates of detected plate box
Merge pull request #49 from holdestmade/main Add Box X and Y centre coordinates attributes for each plate detected
Bugfix
Add On-Premise SDK server
Add On-Premise SDK server courtesy of @gadget-man in #40
Add support for mmc option (paid) to get vehicle orientation
Adds support for setting mmc true in request to receive orientation of vehicle in results.
Doesn't impact default behaviour, but if you've paid for the mmc tier that can return the orientation of the vehicle, setting mmc=true in your config will request the orientation and populate it in the attributes.
Adds watched_plates arg
Adds an optional watched_plates arg which when configured allows individual plates to be watched for. Matching is fuzzy so even if it is not a perfect match (due to poor camera angle for example) the plate can still be matched. The readme gives an example of how a boolean sensor can be created for a watched plate.
Making this 1.0 release as this has all the essential features
arg always_save_latest_jpg is renamed to always_save_latest_file
The optional arg always_save_latest_jpg is now renamed to always_save_latest_file since this is actually a png, and to be consistent with other integrations (deepstack-object, amazon-rekognition)
Small bugfix and pass camera_id to plate recognizer
This release fixes an issue where the state wouldn't update occasionally, and also passes the name of the camera entity to plate recognizer. This means the camera name shows up on the plate recognizer UI:
Adds regions config arg
Add config arg for regions which matches the license plate pattern to a specific region or regions
Adds saving images
See readme