def get_texts_batch(rects, img, Y):
texts = []
features = [
types.Feature(type=enums.Feature.Type.TEXT_DETECTION),
]
requests = []
for i, rect in enumerate(rects):
x, y, w, h = rect
tag = img[int(y):int((y+h)), int(x):int((x+w))]
cv2.imwrite('temp.png', tag)
with open('temp.png', 'rb') as image_file:
imageContext = types.ImageContext(language_hints=["en"])
image = types.Image(content = image_file.read())
request = types.AnnotateImageRequest(image=image, features=features, image_context=imageContext)
requests.append(request)
#print(labelSet[np.argmax(Y[i])])
#texts.append(detect_text('temp.png').strip())
client = vision.ImageAnnotatorClient()
response = client.batch_annotate_images(requests)
for response in response.responses:
if len(response.full_text_annotation.text) > 0:
texts.append(response.full_text_annotation.text.strip())
else:
texts.append("none")
os.remove('temp.png')
return texts
def _annotateBatch(self, batch):
requests = []
for imagePath in batch:
# Loads the image into memory
with io.open(imagePath, 'rb') as image_file:
content = image_file.read()
image = types.Image(content=content)
# https://googlecloudplatform.github.io/google-cloud-python/latest/vision/gapic/v1/types.html#google.cloud.vision_v1.types.AnnotateImageRequest
request = types.AnnotateImageRequest(image=image, features=[
types.Feature(type=FeatureTypes.FACE_DETECTION, max_results=3), # Run face detection.
# types.Feature(type=FeatureTypes.LANDMARK_DETECTION), # Run landmark detection.
# types.Feature(type=FeatureTypes.LOGO_DETECTION), # Run logo detection.
types.Feature(type=FeatureTypes.LABEL_DETECTION, max_results=10), # Run label detection.
# types.Feature(type=FeatureTypes.TEXT_DETECTION), # Run OCR.
# types.Feature(type=FeatureTypes.DOCUMENT_TEXT_DETECTION), # Run dense text document OCR. Takes precedence when both DOCUMENT_TEXT_DETECTION and TEXT_DETECTION are present.
# types.Feature(type=FeatureTypes.SAFE_SEARCH_DETECTION), # Run computer vision models to compute image safe-search properties.
types.Feature(type=FeatureTypes.IMAGE_PROPERTIES), # Compute a set of image properties, such as the image's dominant colors.
types.Feature(type=FeatureTypes.CROP_HINTS, max_results=5), # Run crop hints.
types.Feature(type=FeatureTypes.WEB_DETECTION, max_results=10), # Run web detection.
])
requests.append(request)
response = self.client.batch_annotate_images(requests=requests)
return [
Serializer.responseToJSON(
response.responses[resultCache[imagePath]['index']])
for imagePath in batch
]
def google_request(self):
# Possible features:
# LABEL_DETECTION, FACE_DETECTION, LOGO_DETECTION, TEXT_DETECTION,
# DOCUMENT_TEXT_DETECTION, SAFE_SEARCH_DETECTION, WEB_DETECTION,
# LANDMARK_DETECTION, IMAGE_PROPERTIES
features = [
types.Feature(type=enums.Feature.Type.LABEL_DETECTION),
]
api_requests = []
image = types.Image(content=self.opened_file)
request = types.AnnotateImageRequest(image=image, features=features)
api_requests.append(request)
return self.client.batch_annotate_images(api_requests)
def call_Vision_API(image_binary, requested_features):
'''
Wrapper around the Google Vision API. Enables calling it in one line
with image binary and requested features.
Arguments:
- image_binary: image data
- requested_features: array of strings with features requested.
Returns: AnnotateImageResponse. Check the README for more info on this type.
'''
assert isinstance(image_binary, bytes), 'image_binary should be of class \
bytes.'
assert isinstance(requested_features,
list), 'requested_features should be \
a list.'
# load the image into compatible type
image = types.Image(content=image_binary)
# Instantiates a client
client = vision.ImageAnnotatorClient()
all_features = []
# Attach all the features caller requested
for request in requested_features:
if request not in labels:
print(request, "is not a valid feature.")
# append feature to the list of features to be requested
feature_type = labels[request]
feature = types.Feature(type=feature_type, max_results=10)
all_features.append(feature)
# create the request
request = types.AnnotateImageRequest(image=image, \
features=all_features)
# call the image annotation
result = client.annotate_image(request)
return result
def detect_text(images_array):
"""Detects text in the file."""
from google.cloud import vision
client = vision.ImageAnnotatorClient()
# with io.open(path, 'rb') as image_file:
# content = image_file.read()
features = [
types.Feature(type=enums.Feature.Type.TEXT_DETECTION)
]
requests = []
for filename in images_array:
with open(filename, 'rb') as image_file:
image = types.Image(
content=image_file.read())
request = types.AnnotateImageRequest(
image=image, features=features)
requests.append(request)
response_vision = client.batch_annotate_images(requests)
# response = client.text_detection(image=image)
# texts = response.text_annotations
# print('Texts:')
if response_vision:
return response_vision
# for text in texts:
# print('\n"{}"'.format(text.description))
# vertices = (['({},{})'.format(vertex.x, vertex.y)
# for vertex in text.bounding_poly.vertices])
# print('bounds: {}'.format(','.join(vertices)))
if response_vision.error.message:
raise Exception(
'{}\nFor more info on error messages, check: '
'https://cloud.google.com/apis/design/errors'.format(
response_vision.error.message))
def batch_detect_faces_uri(image_uris=[]):
"""Batch detections of faces via iamge uri"""
if not image_uris:
return []
# Instantiates a client
client = vision.ImageAnnotatorClient()
features = [
# types.Feature(type=enums.Feature.Type.LABEL_DETECTION),
types.Feature(type=enums.Feature.Type.FACE_DETECTION),
]
requests = []
for image_uri in image_uris:
image = types.Image()
image.source.image_uri = image_uri
request = types.AnnotateImageRequest(image=image, features=features)
requests.append(request)
response = client.batch_annotate_images(requests)
face_index = 0
faces_batched = []
for _response in response.responses:
item_date = dict()
item_date['face_request'] = face_index
item_date['from_url'] = image_uris[face_index]
item_date['faces_emotions'] = parse_face_emotions(
_response.face_annotations)
faces_batched.append(item_date)
face_index += 1
return faces_batched
from google.protobuf import json_format
from google.cloud import vision as vision
from google.cloud.vision import enums
from google.cloud.vision import types
import pathlib
# サンプルコードのディレクトリから画像のあるディレクトリへのパスを解決する
image_dir = pathlib.Path(__file__).parent.parent.resolve() / 'image_data'
filename = str(image_dir / 'sample.jpg')
client = vision.ImageAnnotatorClient()
image = vision.types.Image()
image.content = open(filename, 'rb').read()
image_context = vision.types.ImageContext(language_hints=['en'])
request = types.AnnotateImageRequest(
image=image,
features=[{
'type': vision.enums.Feature.Type.DOCUMENT_TEXT_DETECTION
}],
image_context=vision.types.ImageContext(language_hints=['en']))
response_from_url = client.annotate_image(request)
print(response_from_url)
from google.cloud import vision
from google.cloud.vision import enums
from google.cloud.vision import types
# Instantiates a client
client = vision.ImageAnnotatorClient()
features = [
types.Feature(type=enums.Feature.Type.LABEL_DETECTION),
types.Feature(type=enums.Feature.Type.FACE_DETECTION),
]
requests = []
# The name of the image file to annotate
file_name = os.path.join(os.path.dirname(__file__), '[IMAGE].jpg')
# Loads the image into memory
with io.open(file_name, 'rb') as image_file:
content = image_file.read()
image = types.Image(content=content)
request = types.AnnotateImageRequest(image=image, features=features)
requests.append(request)
response = client.batch_annotate_images(requests)
for annotation_response in response.responses:
print(annotation_response)
client = vision.ImageAnnotatorClient() #fetching google vision api client
# You can add other Google Vision's features, like face detection, over here.
features = [
types.Feature(type=enums.Feature.Type.TEXT_DETECTION
) # Here, we are only going to use text detection feature.
]
requests = []
# inputting files here, as in forms of filename
for filename in ['1.jpg', '2.png', '3.jpg']:
with io.open(filename, 'rb') as image_file:
content = image_file.read() # reading the image content
image = vision.types.Image(content=content)
request = types.AnnotateImageRequest(
image=image, features=features) #request is a single dictionary
requests.append(request) #making list of requests for batch request
print("numbers of files submitted to API: " + str(len(requests)))
response = client.batch_annotate_images(
requests
) #response from submitting batch request to Google Cloud API Client
# print(response)
for annotation_response in response.responses:
texts = annotation_response.text_annotations
# text annotations is
# annotation response is
df = pd.DataFrame(columns=['locale', 'description'])
# texts is a list. texts is consisted of 1) entire text paragraph, 2) individual lines following afterwards.