Visionary

Python client library for Google Cloud Vision API

https://cloud.google.com/vision

Install

Library can be installed with pip: pip install visionary

Usage

Usage examples:

from visionary import GoogleCloudVision, LabelDetection, LogoDetection

API_KEY = "dummy_api_key"
client = GoogleCloudVision(API_KEY)

# Ask for logo detection, 10 results max
response = client.annotate("dummy.jpg", LogoDetection())

# Ask for label detection, 3 results max
response = client.annotate("dummy.jpg", LabelDetection(3))

Detection params can be set explicitly:

# 5 results max for logo detection and only one result for label detection
response = client.annotate("dummy.jpg", LogoDetection(5), LabelDetection(1))

First param can be file object:

uploaded_file = open("dummy.jpg")
...
# somewhere later
response = client.annotate(uploaded_file)

Or URL:

response = client.annotate("http://google.com/dummy.jpg")

Client supports multiple images in single annotate call

response = client.annotate(
    ("dummy.jpg"),
    ("http://google.com/dummy.jpg"),
    ("dummy.jpg", LogoDetection(1)),
)

Dealing with response:

if response.ok:
    for resp in response.responses:
        for i in resp.logo_annotations:
            print(i.description)
else:
    print(response.error['code'], response.error['message'], response.error['status'])