core/homeassistant/components/image_processing/seven_segments.py

123 lines
3.9 KiB
Python

"""
Local optical character recognition processing of seven segements displays.
For more details about this platform, please refer to the documentation at
https://home-assistant.io/components/image_processing.seven_segments/
"""
import asyncio
import logging
import io
import os
import voluptuous as vol
import homeassistant.helpers.config_validation as cv
from homeassistant.core import split_entity_id
from homeassistant.components.image_processing import (
PLATFORM_SCHEMA, ImageProcessingEntity, CONF_SOURCE, CONF_ENTITY_ID,
CONF_NAME)
_LOGGER = logging.getLogger(__name__)
CONF_DIGITS = 'digits'
CONF_EXTRA_ARGUMENTS = 'extra_arguments'
CONF_HEIGHT = 'height'
CONF_ROTATE = 'rotate'
CONF_SSOCR_BIN = 'ssocr_bin'
CONF_THRESHOLD = 'threshold'
CONF_WIDTH = 'width'
CONF_X_POS = 'x_position'
CONF_Y_POS = 'y_position'
DEFAULT_BINARY = 'ssocr'
PLATFORM_SCHEMA = PLATFORM_SCHEMA.extend({
vol.Optional(CONF_EXTRA_ARGUMENTS, default=''): cv.string,
vol.Optional(CONF_DIGITS, default=-1): cv.positive_int,
vol.Optional(CONF_HEIGHT, default=0): cv.positive_int,
vol.Optional(CONF_SSOCR_BIN, default=DEFAULT_BINARY): cv.string,
vol.Optional(CONF_THRESHOLD, default=0): cv.positive_int,
vol.Optional(CONF_ROTATE, default=0): cv.positive_int,
vol.Optional(CONF_WIDTH, default=0): cv.positive_int,
vol.Optional(CONF_X_POS, default=0): cv.string,
vol.Optional(CONF_Y_POS, default=0): cv.positive_int,
})
@asyncio.coroutine
def async_setup_platform(hass, config, async_add_devices, discovery_info=None):
"""Set up the Seven segments OCR platform."""
entities = []
for camera in config[CONF_SOURCE]:
entities.append(ImageProcessingSsocr(
hass, camera[CONF_ENTITY_ID], config, camera.get(CONF_NAME)
))
async_add_devices(entities)
class ImageProcessingSsocr(ImageProcessingEntity):
"""Representation of the seven segments OCR image processing entity."""
def __init__(self, hass, camera_entity, config, name):
"""Initialize seven segments processing."""
self.hass = hass
self._camera_entity = camera_entity
if name:
self._name = name
else:
self._name = "SevenSegement OCR {0}".format(
split_entity_id(camera_entity)[1])
self._state = None
self.filepath = os.path.join(self.hass.config.config_dir, 'ocr.png')
crop = ['crop', str(config[CONF_X_POS]), str(config[CONF_Y_POS]),
str(config[CONF_WIDTH]), str(config[CONF_HEIGHT])]
digits = ['-d', str(config[CONF_DIGITS])]
rotate = ['rotate', str(config[CONF_ROTATE])]
threshold = ['-t', str(config[CONF_THRESHOLD])]
extra_arguments = config[CONF_EXTRA_ARGUMENTS].split(' ')
self._command = [config[CONF_SSOCR_BIN]] + crop + digits + threshold +\
rotate + extra_arguments
self._command.append(self.filepath)
@property
def device_class(self):
"""Return the class of this device, from component DEVICE_CLASSES."""
return 'ocr'
@property
def camera_entity(self):
"""Return camera entity id from process pictures."""
return self._camera_entity
@property
def name(self):
"""Return the name of the image processor."""
return self._name
@property
def state(self):
"""Return the state of the entity."""
return self._state
def process_image(self, image):
"""Process the image."""
from PIL import Image
import subprocess
stream = io.BytesIO(image)
img = Image.open(stream)
img.save(self.filepath, 'png')
ocr = subprocess.Popen(
self._command, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
out = ocr.communicate()
if out[0] != b'':
self._state = out[0].strip().decode('utf-8')
else:
self._state = None
_LOGGER.warning(
"Unable to detect value: %s", out[1].strip().decode('utf-8'))