robmarkcole / hass-sighthound Goto Github PK

Firstly, thank you for an awesome add to Home Assistant. I've been playing with this, and thoroughly enjoy it. I have a suggestion for an improvement. As someone who has high traffic in front of their house, it would be interesting to see statistics on vehicles that pass by. For example, right now in my events I can see the following data:

{"entity_id": "image_processing.sighthound_garage_2", "plate": "JDJ7182", "vehicle_type": "car", "make": "Mercedes-Benz", "model": "CLS-Class", "color": "black", "region": "Mississippi", "bounding_box": [0.13503, 0.20703, 0.47454, 0.47266]}

It would be great to gather statistics on how many Ford, Mercedes-Benz, etc... the system recognizes over time. The same for color, region, model, etc...

At this time, the only method I can identify is a database query.

Do robmarkcole/HASS-Deepstack-object#261 here

As title

I tried to add this into HACS and it said that this repo is in the blacklist. Was curious if its possible for it to be un-blacklisted

I cannot rely on internet connection in the area where i plan to use this. So is offline motion detection as trigger possible? I thought the desktop apps is also offline?

Add watched_plate config arg which is a list of plates to watch for, with an entity created for each where the state is the time of the last detection of the plate. If a plate with the same number is matched, or similar by at least 4 characters, update this entity.

As of HA 0.107 most of the functionality of this CC will be in the official integration. The only difference being that this CC provides the sighthound.file_saved events (can be reproduced using the folder_watcher integration), and supports face detection. Update the official integration docs and archive this repo

Is it possible to add a configuration option (boolean true/false) to always save the image, even when no vehicle/people is found?
At the moment the image is written only if detections are greater than zero.

I use the generated image as a local_file camera in HA, so it can be useful to use the latest image to show the updated situation (in a picture-glance card).

Hello,
When upgrated to Home Assistant OS 7.6 I see;

ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/topics/dependency-resolution/#dealing-with-dependency-conflicts
WARNING: You are using pip version 22.0.3; however, version 22.0.4 is available.
You should consider upgrading via the '/usr/local/bin/python3 -m pip install --upgrade pip' command.
2022-04-09 14:15:44 ERROR (stream_worker) [homeassistant.components.stream.stream.camera.kamera_dome1] Error from stream worker: Error opening stream (ERRORTYPE_113, Host is unreachable) rtsp://192.168.1.7/ch0_0.h264
2022-04-09 14:15:45 ERROR (SyncWorker_7) [homeassistant.util.package] Unable to install package pillow==7.2.0: ERROR: Cannot install pillow==7.2.0 because these package versions have conflicting dependencies.
ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/topics/dependency-resolution/#dealing-with-dependency-conflicts
WARNING: You are using pip version 22.0.3; however, version 22.0.4 is available.
You should consider upgrading via the '/usr/local/bin/python3 -m pip install --upgrade pip' command.

in logs. Can Ypu please tell me what to do?

As title

I keep getting an error

2018-06-19 09:03:52 ERROR (SyncWorker_9) [custom_components.image_processing.sighthound] sighthound error code 401: {"error":"ERROR_UNAUTHORIZED","reason":"Unauthorized token","reasonCode":40102}

I tried the curl command and get the same error:

arsaboo@aloknuc:~$ curl -X POST https://dev.sighthoundapi.com/v1/detections?type=face,person&faceOption=landmark,gender \
> -H "Content-Type: application/json" \
> -H "X-Access-Token: APIKEY" \
> -d '{"image": "https://images.pexels.com/photos/415829/pexels-photo-415829.jpeg"}'
[1] 31237
-H: command not found
arsaboo@aloknuc:~$ {"error":"ERROR_UNAUTHORIZED","reason":"Required header: X-Access-Token","reasonCode":40101}

Changing the order in the curl command works. So, the following works:

arsaboo@aloknuc:~$ curl -X POST -H "Content-Type: application/json" -H "X-Access-Token: APIKEY" -d '{"image": "https://images.pexels.com/photos/415829/pexels-photo-415829.jpeg"}' https://dev.sighthoundapi.com/v1/detections?type=face,person&faceOption=landmark,gender

My Hikvision ANPR camera throws out lots of ‘defocus‘ events, which aren’t in the list of event constants for pyhik and fill the hass log with errors. It’s one of the Smart Events on the camera, though switching those off in the config UI makes no difference. Possible to add this event name to the list?

I'm using this custom integration through HACS. I'm running Home Assistant on Docker (version 0.105). I'm able to run image_processing.scan successfully and it saves the processed image to my www folder. The person_detected' event fires but the file_saved` event is never fired. So I'm struggling to use this in an automation to send the image in a notification. Let me know what additional information is needed to help debug this error.

How did you get the vechicle make and model to show up? When I run Image processing: scan all I get is the following in attributes

last_vehicle: 2021-05-12_16:46:17
plates: unknown, unknown
unit_of_measurement: vehicle
friendly_name: sighthound_vehicle_driveway

As reported on forums:

File “/config/custom_components/sighthound/image_processing.py”, line 112, in process_image
self.faces = hound.get_faces(detections)
File “/usr/local/lib/python3.7/site-packages/simplehound/core.py”, line 54, in get_faces
for obj in detections[“objects”]:
TypeError: ‘NoneType’ object is not subscriptable

Caused when detections is None and we are calling get_faces(). We need to check and exit on None detections

The following error is caused by status code not being checked:

Update for image_processing.sighthound_porch fails
Traceback (most recent call last):
  File "/srv/homeassistant/lib/python3.6/site-packages/homeassistant/helpers/entity.py", line 197, in async_update_ha_state
    yield from self.async_device_update()
  File "/srv/homeassistant/lib/python3.6/site-packages/homeassistant/helpers/entity.py", line 318, in async_device_update
    yield from self.async_update()
  File "/srv/homeassistant/lib/python3.6/site-packages/homeassistant/components/image_processing/__init__.py", line 145, in async_update
    yield from self.async_process_image(image.content)
  File "/usr/local/lib/python3.6/concurrent/futures/thread.py", line 56, in run
    result = self.fn(*self.args, **self.kwargs)
  File "/home/homeassistant/.homeassistant/custom_components/image_processing/sighthound.py", line 110, in process_image
    self.faces, self.persons = parse_api_response(response)
  File "/home/homeassistant/.homeassistant/custom_components/image_processing/sighthound.py", line 41, in parse_api_response
    for obj in response.json()['objects']:
KeyError: 'objects'

Is there anyway, sorry if i missed it, to get the car color and model detection working?

as title

I have one of my cameras pointed at my driveway. Adding in an additional event for "vehicle detected" -- ideally passing the additional metadata representing the vehicle and license plate -- would allow me to know who is coming and going. We could even trigger automations based on specific license plates (!)

https://www.sighthound.com/docs/cloud/recognition/#vehicle-recognition