Image Intelligence offers a RESTful HTTP API allowing you to integrate intelligent analytics and monitoring to your platform. It uses a blend of artificial and human intelligence to ensure the most accurate results.

If this is the first time you're here, please check out our quickstart guide. It walks you through how to create an account, generate an API accessToken using a clientId/clientSecret and shows you how to make your first API call.

Asynchronous API and Jobs

Our HTTP API operates in batches (i.e. multiple images at a time). Depending on the operation (e.g. object detection, image classification, face recognition), processing each image can take a bit of time and it's due to this, our API needs to be asynchronous.

When a POST request is made to one of our endpoints e.g. /recognize or /detect, a result won't be returned immediately. Instead, we respond a jobStatus and jobId for you to lookup later. We also provide a way to hook onto events (see webhooks).

A Job is an asynchronous operation initiated by you to process a series of images at a later time*. Jobs are the core of the API, allowing you to isolate data between requests, store context, and reference historical results. Almost all endpoints operate on jobs or resources that belong to a job.

*We try to ensure jobs are completed as quickly as possible. Typically this means within a few seconds, usually a few ms. The speed varies between the types of jobs and how many images are sent in each job.

  • /oauth/token: Obtain an accessToken using your application's clientId and clientId
  • /recognize: Detect and recognize objects (e.g. faces) in an image
  • /detect: Performs image classification on common classes such as person and car
  • /match: Find objects that matches any target in list of specified samples
  • /ask: Using Human Intelligence, answers a yes-or-no question from a set of sample images
  • /feedback: Submit issues you find with your results from our API

You can read more about each API endpoint by navigating the sidebar menu.