Analyzing images using Visual Recognition built-in models

IBM Watson Visual Recognition comes with built-in models that you can use to analyze images for scenes, objects, faces, and many other categories without any training.

 

Python notebook example

Here is an example of Python code that uses Visual Recognition built-in models to analyze a sample image. This sample code calls the Visual Recognition API external link and can be run in a notebook in IBM Watson Studio.

!pip install --upgrade "watson-developer-cloud>=1.0,<2.0"
from watson_developer_cloud import VisualRecognitionV3
visual_recognition = VisualRecognitionV3( '2016-05-20', api_key='<your-API-key>' )
image_url = 'https://watson-developer-cloud.github.io/doc-tutorial-downloads/visual-recognition/visual-recognition-food-fruit.png'
import json
parms = json.dumps( { 'url' : image_url, 'classifier_ids' : [ 'food' ] } )
results = visual_recognition.classify( parameters = parms )
print( json.dumps( results['images'][0]['classifiers'][0]['classes'], indent=2 ) )

For information about how to look up your API key, see: Building Visual Recognition apps.

Sample image

Image of a fruit basket

Sample output

[
  {
    "class": "lemon",
    "score": 0.583,
    "type_hierarchy": "/fruit/citrus/lemon"
  },
  {
    "class": "citrus",
    "score": 0.719
  },
  {
    "class": "fruit",
    "score": 0.901
  },
  {
    "class": "apple",
    "score": 0.526,
    "type_hierarchy": "/fruit/accessory fruit/apple"
  },
  {
    "class": "accessory fruit",
    "score": 0.526
  },
  {
    "class": "orange",
    "score": 0.518,
    "type_hierarchy": "/fruit/citrus/orange"
  },
  {
    "class": "banana",
    "score": 0.5,
    "type_hierarchy": "/fruit/banana"
  }
]

 

See also