MMALSharp - C# wrapper to Broadcom's MMAL and API to the Raspberry Pi camera

If you like this project, please support it by giving it a star!

MMALSharp is a C# wrapper around the MMAL library designed by Broadcom. It exposes many elements of MMAL and in addition provides an easy to use, asynchronous API to the Raspberry Pi Camera Module. The library targets .NET Standard 2.0 and is compatible with Mono 5.4/.NET Core 2.0 or greater platforms.

Installation

MMALSharp NuGet package:

PM> Install-Package MMALSharp

MMALSharp.FFmpeg NuGet package:

PM> Install-Package MMALSharp.FFmpeg

Pre-release builds can be found on MyGet:

Logging configuration

For v0.6, MMALSharp now uses Microsoft.Extensions.Logging.Abstractions to provide package agnostic logging. If you want to enable logging, you must provide the ILoggerFactory instance your client application is using. For .NET Core applications, this will typically be done during dependency injection configuration. For more information, please see here.

Below is an example on how to configure NLog in a .NET Core 3.0+ console app. Note: the ILoggerFactory instance should be set before carrying out any MMALSharp specific operations:

var loggerFactory = LoggerFactory.Create(builder =>
{
	builder                
	.ClearProviders()
	.SetMinimumLevel(LogLevel.Trace)                
	.AddNLog("NLog.config");
});

MMALLog.LoggerFactory = loggerFactory;

Also see here for full NLog integration instructions.

Basic Examples

Take a JPEG image using YUV420 encoding:

public void TakePicture()
{
    // Singleton initialized lazily. Reference once in your application.
    MMALCamera cam = MMALCamera.Instance;

    using (var imgCaptureHandler = new ImageStreamCaptureHandler("/home/pi/images/", "jpg"))        
    {            
        await cam.TakePicture(imgCaptureHandler, MMALEncoding.JPEG, MMALEncoding.I420);
    }
    
    // Cleanup disposes all unmanaged resources and unloads Broadcom library. To be called when no more processing is to be done
    // on the camera.
    cam.Cleanup();
}

Take a H.264 video using YUV420 encoding at 30 fps:

public void TakeVideo()
{
    // Singleton initialized lazily. Reference once in your application.
    MMALCamera cam = MMALCamera.Instance;

    using (var vidCaptureHandler = new VideoStreamCaptureHandler("/home/pi/videos/", "avi"))        
    {    
        var cts = new CancellationTokenSource(TimeSpan.FromMinutes(3));
                
        await cam.TakeVideo(vidCaptureHandler, cts.Token);
    }   

    // Cleanup disposes all unmanaged resources and unloads Broadcom library. To be called when no more processing is to be done
    // on the camera.
    cam.Cleanup();
}

Basic Configuration

Below are common global configuration properties which you may want to edit:

MMALCameraConfig.Resolution = new Resolution(640, 480); // Set to 640 x 480. Default is 1280 x 720.
MMALCameraConfig.Framerate = new MMAL_RATIONAL_T(20, 1); // Set to 20fps. Default is 30fps.
MMALCameraConfig.ShutterSpeed = 2000000; // Set to 2s exposure time. Default is 0 (auto).
MMALCameraConfig.ISO = 400; // Set ISO to 400. Default is 0 (auto).

Performance

The performance of MMALSharp is very good due to being a wrapper around the native MMAL library with minimal additional operations being carried out whilst the camera is in use. BenchmarkDotNet reports a mean time of 527.0ms to capture a JPEG still at 640 x 480 resolution following configuration and warmup of the camera, with even faster still capture performance being gained if you capture from the camera's video port instead.

SD Card IO performance - Having a high performance SD card is crucial to get the most out of using this library, especially when saving to multiple FileStreams with the splitter component. You may find when outputting to multiple FileStreams that the SD Card IO is a bottleneck and using a faster storage medium would be beneficial such as a ramdisk mount. On the Raspberry Pi 4, using an external USB 3 Hard Drive would also be a great option for the best performance possible.

Documentation

For full installation instructions for Mono and .NET Core, including configuration and examples - please visit the Wiki.

License

MIT license

Copyright (c) 2016-2020 Ian Auty and contributors.

Raspberry Pi is a trademark of the Raspberry Pi Foundation

Contributors

I want to say a big thank you to those of you who have helped develop MMALSharp over the years, your contributions are most appreciated. In addition, I'd like to say thanks to Dave Jones @waveform80 for your work on picamera which gave me the inspiration to start this project.