Getting Started

1. What is OpenCV.NET

OpenCV.NET is a .NET interface for the popular OpenCV computer vision and image processing library. These bindings were developed from scratch over several years and include many features that have been missing from other attempts, specifically:

  • Cross-platform compatibility with a .NET standard API (only P/Invoke is used).
  • Very close to 100% support of the C API (including undocumented functions).
  • Full documentation/intellisense for all supported methods, classes and enum members (arduously mined from many different sources).
  • Support for exceptions.
  • Garbage-collector friendly:
    • GC is aware of the actual memory used by matrices and images.
    • Pointers are guaranteed to survive function calls even if concurrent garbage collection happens.
  • Object model following C++ API as much as possible (towards future increased support of C++ API).
  • Permissive MIT license.
Warning

The bindings currently target OpenCV 2.4.8. OpenCV has since moved towards a pure C++ API and has subsequently deprecated its C interface, which means we cannot easily access new features. We are currently exploring different alternatives to generate a comprehensive high-quality wrapper for newer versions of OpenCV, but for now OpenCV.Net will keep targeting the 2.4.x series.

2. Installing OpenCV.NET

The easiest way to start using OpenCV.NET in your projects is to install the NuGet package from one of the following sources:

  • .NET Command-line: dotnet add package OpenCV.NET.
  • Powershell: From the NuGet Package Manager Console, type Install-Package OpenCV.Net.
  • PackageReference: Add the line <PackageReference Include="OpenCV.Net" Version="3.3.1" /> to your .csproj file.
Important

Prerequisites The Visual C++ Redistributable for Visual Studio 2012 may be required on Windows systems to correcty load native dependencies.