Using Windows 10 Built-in OCR With C#
As you know, from beginning Windows 10, a client-based OCR engine was included. This engine was a product of Microsoft Research’s project ‘Oxford,’ and this project related to Microsoft Azure’s Cognitive Service.
There are plenty of options to choose what OCR engine and how to use it. In my case, I want to select and test the Desktop built-in OCR engine, which does not make any additional cost, so I focused on it solely.
Requirements
Windows 10’s OCR is available with the Universal Windows Platform (UWP) API. But most UWP API seems like available only in UWP application, not for Desktop App. Luckily, OCR API is the exception, so you can easily use the API. Before following this article, you need to install below dependencies.
- Windows 10 SDK (Matching it against your current Windows 10 build)
- The latest version of Visual Studio (I tested it initially on Visual Studio 2010) or LINQPAD 5
For simplicity, I’m using LINQPAD 5.
