Skip to content

vision-usage.rst

Latest commit

 

History

History
233 lines (172 loc) · 5.64 KB

vision-usage.rst

File metadata and controls

233 lines (172 loc) · 5.64 KB

Using the Vision API

Authentication and Configuration

  • For an overview of authentication in gcloud-python, see :doc:`gcloud-auth`.

  • In addition to any authentication configuration, you should also set the :envvar:`GCLOUD_PROJECT` environment variable for the project you'd like to interact with. If the GCLOUD_PROJECT environment variable is not present, the project ID from JSON file credentials is used.

    If you are using Google App Engine or Google Compute Engine this will be detected automatically.

  • After configuring your environment, create a :class:`Client <gcloud.vision.client.Client>`

>>> from gcloud import vision
>>> client = vision.Client()

or pass in credentials and project explicitly

>>> from gcloud import vision
>>> client = vision.Client(project='my-project', credentials=creds)

Annotating an Image

Annotate a single image

>>> from gcloud import vision
>>> client = vision.Client()
>>> image = client.image('./image.png')
>>> faces = image.detect_faces(limit=10)

Annotate multiple images

>>> first_image = client.image('./image.jpg')
>>> second_image = client.image('gs://my-storage-bucket/image2.jpg')
>>> with client.batch():
...     labels = first_image.detect_labels()
...     faces = second_image.detect_faces(limit=10)

or

>>> images = []
>>> images.append(client.image('./image.jpg'))
>>> images.append(client.image('gs://my-storage-bucket/image2.jpg'))
>>> faces = client.detect_faces_multi(images, limit=10)

No results returned

Failing annotations return no results for the feature type requested.

>>> from gcloud import vision
>>> client = vision.Client()
>>> image = client.image('./image.jpg')
>>> logos = image.detect_logos(limit=10)
>>> logos
[]

Manual Detection

You can call the detection method manually.

>>> from gcloud import vision
>>> client = vision.Client()
>>> image = client.image('gs://my-test-bucket/image.jpg')
>>> faces = image.detect(type=vision.FACE_DETECTION, limit=10)

Face Detection

Detecting a face or faces in an image. For a list of the possible facial landmarks see: https://cloud.google.com/vision/reference/rest/v1/images/annotate#type_1

>>> from gcloud import vision
>>> client = vision.Client()
>>> image = client.image('./image.jpg')
>>> faces = image.detect_faces(limit=10)
>>> faces[0].landmarks[0].type
'LEFT_EYE'
>>> faces[0].landmarks[0].position.x
1301.2404
>>> faces[0].detection_confidence
0.9863683
>>> faces[0].joy_likelihood
0.54453093
>>> faces[0].anger_likelihood
0.02545464

Label Detection

Image labels are a way to help categorize the contents of an image. If you have an image with a car, person and a dog it, label detection will attempt to identify those objects.

>>> from gcloud import vision
>>> client = vision.Client()
>>> image = client.image('./image.jpg')
>>> labels = image.detect_labels(limit=3)
>>> labels[0].description
'automobile'
>>> labels[0].score
0.9863683

Landmark Detection

The API will attemtp to detect landmarks such as Mount Rushmore and the Sydney Opera House. The API will also provide their known geographical locations if available.

>>> from gcloud import vision
>>> client = vision.Client()
>>> image = client.image('./image.jpg')
>>> landmarks = image.detect_landmarks()
>>> landmarks[0].description
'Sydney Opera House'
>>> landmarks[0].locations[0].latitude
-33.857123
>>> landmarks[0].locations[0].longitude
151.213921
>>> landmarks[0].bounding_poly.vertices[0].x
78
>>> landmarks[0].bounding_poly.vertices[0].y
162

Logo Detection

Google Vision can also attempt to detect company and brand logos in images.

>>> from gcloud import vision
>>> client = vision.Client()
>>> image = client.image('./image.jpg')
>>> logos = image.detect_logos(limit=1)
>>> results.logos[0].description
'Google'
>>> logos[0].score
0.9795432
>>> logos[0].bounding_poly.vertices[0].x
78
>>> logos[0].bounding_poly.vertices[0].y
62

Safe Search Detection

Detecting safe search properties of an image.

>>> from gcloud import vision
>>> client = vision.Client()
>>> image = client.image('./image.jpg')
>>> safe_search = image.detect_safe_search()
>>> safe_search.adult
'VERY_UNLIKELY'
>>> safe_search.medical
'UNLIKELY'

Text Detection

Detecting text with ORC from an image.

>>> from gcloud import vision
>>> client = vision.Client()
>>> image = client.image('./image.jpg')
>>> text = image.detect_text()
>>> text.locale
'en'
>>> text.description
'the full text of the image.'

Image Properties

Detecting image color properties.

>>> from gcloud import vision
>>> client = vision.Client()
>>> image = client.image('./image.jpg')
>>> colors = image.detect_properties()
>>> colors[0].red
244
>>> colors[0].blue
134
>>> colors[0].score
0.65519291
>>> colors[0].pixel_fraction
0.758658