The installation process does not bring any kind of unpleasant surprises and it over in a few seconds, while the interface you come across presents a design which can be described as minimal and modern. Hassle-free setup and simple-to-handle UI Plot_person: 'python /path/to/plot.Sighthound Video, formerly known as Vitamin D Video, is a software tool that was developed specifically to aid individuals in setting up a surveillance system with motion detection capabilities and saving videos to the hard drive. The shell_command could look something like this: If you're running in Docker you'll have to run this command inside the docker container. But first you'll need to install the gmplot package using pip install gmplot. Similar to the export script, this script can be called using a shell_command. Python plot.py -p PERSON_NAME Use with Home Assistant This will generate a HTML file that can be displayed in a browser. The plog script can be used to plot location data in a.

You can now create an automation that calls this script whenever the device_tracker changes, or create a new flow in Node-Red. Note that the device tracker needs to include latitude and longitude attributes. This can then be displayed using a local_image camera. This will automatically save an image of the camera with detected objects drawn over it. Note: An easier way to get this done is by using the file_out parameter of the Tensorflow component. This Tensorflow camera component will update automatically, drawing the detected people and faces over the latest snapshot image and displaying that. Whenever a new snapshot is created, the image_processing.scan service for Tensorflow is called. A local_file camera is set up displaying this latest.jpg image.

RTSP camera > local_file camera > Tensorflow image_processing > Tensorflow cameraĪ snapshot is taken using the camera.snapshot service whenever motion is detected and the image is saved under /images/latest.jpg. My current image_processing flow is the following: Since image_processing components can take a while to process an image, the best way to get this up and running is by using a local_file camera. This component will draw detected objects, using different colors for different object types. Processor: image_processing.sighthound_livingroom_latestĪ camera component that displays objects detected by the Tensorflow component.

Sighthound camera feed skyipcam1620