Building an OCR service

In this section, we will be using some of the AI services using C# as well as PowerShell to show their usage using the HTTP endpoint directly. The next section will concentrate on doing the same using a .NET SDK.

Before getting into building a project using Cognitive Services, the first step is to provision the API itself.

Optical character recognition is available as a Vision API and can be provisioned using the Azure portal, as shown next. Create a vision API by navigating to Cognitive Services > Compute Vision > Create, as shown in Figure 19.1:

Figure 19.1: Create a Vision API

Once the API is provisioned, the overview page provides all the details for consuming the API. It provides the base URL and the key information. Make a note of the key as it will be used later:

Figure 19.2: Overview page

It also provides an API console to quickly test the API. Clicking on it opens a new window...