opencvsharp | OpenCV wrapper for .NET | Build Tool library

When capturing via OpenCvSharp, 640x480 is the default resolution.

You must set the desired resolution before the device is opened (which is done implicitly when you grab frames) e.g.:

You would create a new bitmap, at least as large as the bounding box of your polygon. Create a graphics object from this new bitmap. You can then draw the polygon to this bitmap, using the original image as a texture brush. Note that you might need to apply transform matrix to translate from the full image coordinates to the cropped image coordinates.

Note that it looks like you have radiological images. These are typically 16 bit images, so they will need to be converted to 8bit mono, or 24bit RGB before they can be used. This should already be done in the drawing code if you have access to the source. Or you can do it yourself.

OpenCvSharp4 supports .NET5 (and also .NET6 preview :-)).

1. Create an empty "Windows Forms App" project form Visual Studio. (Obviously don't use .NET Framework...)

2. Add nuget package OpenCvSharp4.Windows. Below my csproj file.

I found them, on the opencv_contrib GitHub page.

You can find both YML files (the model and the range files) here.

I found a solution myself after a lot more playing around.

A single thread to capture the frames and write them was not enough. I now created two threads, one that captures the frames from the camera and one that writes them. To have the correct timing in the resulting file the delay the write creates has to be taken into account.

I ended up with the following two functions that run in separate threads:

Good to see you resolved the issue.

Regarding your update with the extra code -- I'd be really cautious about using the Assembly GetTypes() and the Activator for it. It's always better to create instances of types you have static access to rather than going through reflection or breaking interface rules. I believe GetTypes() returns new type instances, and since Type .Equals/== only checks reference of the underlying system type, then a new instance will not be equal.

I wonder if your problem is the two Box types look the same but the instances are different, then the runtime is not able to convert the two.

I think that correct way should be this:

This may not be exactly what you need, but it should give you a basic idea of how to create different binarized bitmaps from the same, unmodified raw pixel buffer.

In a first step, it creates a 8-bits-per-pixel grayscale BitmapSource from the original BitmapSource, and copies its buffer to a byte array. Each byte in the buffer is a gray value in the range 0 to 255. The trick is to reuse this as an index into a color palette with 256 entries.

Now each time the threshold value (i.e. the Value of a Slider with range 0 to 255) changes, a color palette with 256 entries is created, but it only contains Black for indices below the threshold, and Whitefor indices above.

These packages are incompatible with .NET Core 3.1.

• WpfWebcamControl (>= .NET Framework 4.6)
• WebcamCapturer.Core (>= .NET Framework 4.7.2)

For OpenCV, you might be using the wrong package, as it is indeed compatible with .NET Core 3.1.

• OpenCVSharp (.NET Framework)
• OpenCVSharp4 (>= .NET Framework 4.6.2, >= .NET Standard 2.0, >= .NET Core 2.1)
• For more info and packages, see GitHub and a short tutorial here.

Please note that the OpenCVSharp4 package is the core package, depending on your use-case, you might instead use OpenCvSharp4.Windows and optionally OpenCvSharp4.WpfExtensions. Please refer to the project page GitHub and the Wiki for more information.