A custom camera component that allows you to see the detected regions from the OpenCV image processor in Home-Assistant. Add this to your configuration directory: {CONFIG_DIRECTORY}/custom_components/opencv.py

example.yaml

camera:
  - platform: opencv
    camera: camera.front_door
    processor: image_processing.front_door_opencv_faces
    # Optional Parameters
    name: OpenCV Camera

opencv.py

"""
Camera that highlights the detected regions from the opencv image processor
"""
import asyncio
import io
import logging
import os

import voluptuous as vol

from homeassistant.const import CONF_NAME
from homeassistant.components.camera import Camera, PLATFORM_SCHEMA
from homeassistant.helpers import config_validation as cv
from homeassistant.components.image_processing.opencv import ATTR_MATCHES
from homeassistant.loader import get_component

_LOGGER = logging.getLogger(__name__)

ATTR_CAMERA = 'camera_entity'
ATTR_PROCESSOR = 'processor_entity'

CONF_CAMERA = 'camera'
CONF_COLOR = 'color'
CONF_PROCESSOR = 'processor'
CONF_CLASSIFIER = 'classifier'

DEFAULT_COLOR = (255, 255, 0)
DEFAULT_NAME = 'OpenCV'

PLATFORM_SCHEMA = PLATFORM_SCHEMA.extend({
    vol.Optional(CONF_NAME, default=DEFAULT_NAME): cv.string,
    vol.Required(CONF_CAMERA): cv.entity_id,
    vol.Required(CONF_PROCESSOR): cv.entity_id,
    vol.Optional(CONF_CLASSIFIER, default=None): cv.string,
    vol.Optional(CONF_COLOR, default=DEFAULT_COLOR): (int, int, int)
})


def setup_platform(hass, config, add_devices, discovery_info=None):
    """Set up the OpenCV camera platform."""
    add_devices([OpenCVCamera(hass, config.get(CONF_NAME, DEFAULT_NAME),
                             config[CONF_CAMERA], config[CONF_PROCESSOR],
                             config[CONF_CLASSIFIER], config[CONF_COLOR])])


class OpenCVCamera(Camera):
    """Visual representation of opencv matched regions."""
    def __init__(self, hass, name, camera, processor, classifier, color):
        """Initialize the opencv camera."""
        super().__init__()
        self._hass = hass
        self._camera = camera
        self._processor = processor
        self._color = color
        self._name = name
        self._classifier = classifier

    @property
    def name(self):
        """Return the name of this camera."""
        return self._name

    @property
    def state_attributes(self):
        """Return the device state attributes."""
        return {
            ATTR_CAMERA: self._camera,
            ATTR_PROCESSOR: self._processor
        }

    @asyncio.coroutine
    def async_camera_image(self):
        """Return the camera image still."""
        from PIL import Image, ImageDraw

        camera = get_component('camera')
        image = None
        processor = self._hass.states.get(self._processor)

        try:
            image = yield from camera.async_get_image(
                self._hass, self._camera, timeout=2)

        except HomeAssistantError as err:
            _LOGGER.error("Error on receive image from entity: %s", err)
            return

        matches = processor.attributes.get(ATTR_MATCHES)
        regions = []
        if self._classifier is None:
            for key, value in matches.items():
                for region in value:
                    regions.append(region)
        elif self._classifier in matches:
            for region in matches[self._classifier]:
                regions.append(region)
        else:
            _LOGGER.error("Cannot locate classifier %s", self._classifier)
            return
        if len(regions) == 0:
            return image


        stream = io.BytesIO(image)
        im = Image.open(stream)
        annotated_image = ImageDraw.Draw(im)
        for region in regions:
            x0 = region[0]
            y0 = region[1]
            x1 = x0 + region[2]
            y1 = y0 + region[2]
            annotated_image.rectangle([x0, y0, x1, y1],
                                      outline=self._color)

        image_bytes = io.BytesIO()
        im.save(image_bytes, format='PNG')
        return image_bytes.getvalue()