Nuxeo integrates with Google Cloud Vision API for automated image recognition

March 31, 2016  - By 0 Comments
Image objects/themes (“horizon”, “nature”) and landmarks (“savannah”) found by Google Cloud Vision service automatically added as tags.

Image objects/themes (“horizon”, “nature”) and landmarks (“savannah”) found by Google Cloud Vision service automatically added as tags.

The Nuxeo Platform is integrating with the Google Cloud Vision API, a new service providing information and insights automatically detected within images. The plugin is now available to download directly from the Nuxeo Marketplace, which offers optional packages to easily add new features to the Nuxeo Platform.

The Nuxeo Platform enables organizations to manage complex digital content at massive scale.

Using the plugin, the Nuxeo Platform exposes a powerful set of configurable image analysis features that can be applied upon file import, within a given workflow/business process, and/or on demand after a specific user action.

Nuxeo also applies Google Cloud Vision to videos using images (video frames) captured by the Nuxeo storyboard, identifying landmarks (“Eiffel Tower”) and objects/themes (“human action”, “mobile device”) saved as tags.

Nuxeo also applies Google Cloud Vision to videos using images (video frames) captured by the Nuxeo storyboard, identifying landmarks (“Eiffel Tower”) and objects/themes (“human action”, “mobile device”) saved as tags.

In addition to processing any image attached to content objects, the Nuxeo Platform also applies the capabilities to video, using images extracted from the video storyboard.

Image features detected by the Google Cloud Vision service are available for use with Nuxeo process automation and business logic, and can also be stored as metadata and tags on the content object, including:

  • Image labeling based on detection of common objects, landmarks, and/or brand logos
Faces, if any, detected in the image, including simple sentiment recognition (e.g., “joy,” “sorrow”)
  • Text found within the image, extracted via OCR
  • Safe search detection of different types of inappropriate content
  • Other general image attributes, such as the dominant color

Additionally, the Nuxeo plugin enables developers to specify business logic for customized operations with the Google Cloud Vision service, such as invoking workflows based on the presence of automatically added image metadata.

“The Google Vision team has done exceptional work to develop this fast, reliable and incredibly valuable service,” said Eric Barroca, CEO at Nuxeo. “When the Google Cloud Vision API was placed in public beta last month, we recognized this was game-changing technology for media and digital asset management applications. We immediately went to work integrating this service into the Nuxeo Platform, and we’re proud to be among the first content management vendors to do so.”

Barroca added, “We strongly believe our integration with Google Vision greatly extends the strategic value of the Nuxeo Platform for our customers. By automatically adding a wealth of new, actionable metadata describing content within each image, Nuxeo enables a whole new level of image-based enterprise search, workflows and automation.”

The Nuxeo plugin, named Nuxeo Vision, has been developed to immediately support the Google Cloud Vision service, as well as other image recognition engines in the future.

The Google Cloud Vision API enables developers to understand the content of an image by encapsulating powerful machine learning models in a REST API.

This article is tagged with , , , and posted in Featured Stories, GIS News, GIS Software, Technology
Tracy Cozzens

About the Author:

Tracy Cozzens has served as managing editor of GPS World magazine since 2006. She also is editor of GPS World’s newsletters and the sister website Geospatial Solutions. She has worked in government, for non-profits, and in corporate communications, editing a variety of publications for audiences ranging from federal government contractors to teachers.

Post a Comment