Community

Elevate Watson Image Recognition with Bluemix Services

Share this post:

To see what Watson Image Recognition with Bluemix infrastructure and services can do together, you need a problem that seems insurmountable.

Consider this dilemma:

So far, 20 of your homeowners insurance policy-holders have filed claims for roof damage sustained in a recent hailstorm. More claims are expected.

Working the usual way, site inspections and documentation of damage will have high costs and take weeks or months to complete. Your business will be lucky if customer dissatisfaction is its only form of collateral damage.

A different hail–for an IBM Bluemix solution

Using a DJI Phantom Professional drone and leveraging Bluemix instructure and services, IBM Developer Advocate Andy Trice creates a Bluemix application to generate a detailed property assessment–including images, a full photogrammatic 3D model of the property and damage, and a reliable claim estimate–within hours instead of days or weeks.

How this solution puts Watson Image Recognition to work on Bluemix

First, the user of the property assessment app performs the following actions:

  • Initiates a property damage assessment in a web interface.
  • Loads images into Bluemix Object Storage captured from a drone that takes high resolution photos of each property—usually generating 50-100 12 megapixel images.

Then the assessment app automatically initiates the following:

  • Creates metadata for the media files in a Cloudant database associated with the assessment.
  • Leverages a node.js app to break up the large drone images into smaller tiles that are sent into the Watson Image Recognition service for auto-identification of damage in each tile. This involves previously defining customer classifiers that teaches Watson how to recognize areas of an image as hail-damaged. Simply put, seeding Watson with images of undamaged shingle roofs becomes a baseline against which to identify areas of damage.
  • Generates a photogrammetric 3D model of the property based on the drone images–which have embedded x, y, z geospatial coordinates from which the assessment app represents shapes as point clouds whose dimensions are reliably measurable in configurable units (meters by default).

Finally, the user performs these steps to finish the assessment:

  • Overlays Watson-analyzed damage onto captured images.
  • Measures areas of roof damage in the 3D model.
  • Estimates claims based on the measurements of damage, current market costs for repairs, and homeowners policy coverage.

With this new solution, you can lower costs, accelerate assessment of damaged areas, and file claims in only hours – improving your overall customer satisfaction.

Check out Andy’s step by step video demostration. Andy hasn’t yet given the app a name, but what would you call it?

Learn more about the Watson Visual Recognition with Bluemix Services and if you’re ready, try Bluemix out for yourself.

Senior Content Strategist, WW Marketing

More Community stories
April 29, 2019

Transforming Customer Experiences with AI Services (Part 1)

This is an experience from a recent customer engagement on transcribing customer conversations using IBM Watson AI services.

Continue reading

April 26, 2019

Updated Tutorial: Database-Driven Chatbot

The tutorial on how to build a database-driven chatbot has been updated. It's now simpler to deploy and and offers more options—Slack, Facebook Messenger, Wordpress, and more.

Continue reading

April 23, 2019

Announcing the Deprecation of the Watson Machine Learning JSON Token Authentication Service

We’d like to inform you about the deprecation of the Watson Machine Learning JSON Token Authentication service. This method of authentication will be retired on May 30, 2019.

Continue reading