Emgu.CV.BinaryMatching 1.0.1

Emgu.CV.BinaryMatching

Emgu.CV.BinaryMatching is a high-performance C# library built on top of EmguCV, designed to perform template matching on binary images with rotation invariance. This library offers exceptional speed, making it suitable for real-time applications and capable of detecting multiple objects in an image.

Features

• Binary Image Template Matching: Specifically optimized for binary (black and white) images.
• Rotation Invariance: Detects objects regardless of their rotation angle.
• High Speed: Efficient implementation allows real-time processing.
• Multiple Object Detection: Capable of identifying more than one matching object in the image.

Requirements

• .NET Framework 4.6.1+
• Emgu.CV 4.9.0.5494+
• Emgu.CV.Bitmap 4.9.0.5494+
• Emgu.CV.runtime.windows 4.9.0.5494+

Installation

You can install Emgu.CV.BinaryMatching via NuGet:

Install-Package Emgu.CV.BinaryMatching -Version 1.0.1

Usage

Please set the build architecture to either "x86" or "x64". Do not set the architecture to "Any CPU". Here's a simple example demonstrating how to use the Emgu.CV.BinaryMatching library:

using Emgu.CV;
using Emgu.CV.BinaryMatching;
using Emgu.CV.CvEnum;
using System.Linq;

public class Example
{
	public void DetectObject()
	{
		Mat src = CvInvoke.Imread("E:\\src.jpg", ImreadModes.Grayscale);
		Mat templ = CvInvoke.Imread("E:\\template.jpg", ImreadModes.Grayscale);

		Mat binarySrc = new Mat();
		Mat binaryTempl = new Mat();

		// Adjust the parameters appropriately.
		CvInvoke.Threshold(src, binarySrc, 70, 255, ThresholdType.BinaryInv);
		CvInvoke.Threshold(templ, binaryTempl, 70, 255, ThresholdType.BinaryInv);

		CvInvoke.Imshow("binary src", binarySrc);
		CvInvoke.Imshow("binary templ", binaryTempl);

		/* Find one matched object
		* "templScore" is MatchTemplate score
		* "shapeScore" is MatShapes score*/
		MatchedObject obj = FlatMatcher.FindOne(binarySrc, binaryTempl, templScore: 0.9, shapeScore: 0.1);

		/* Or use Find if you want to find a maximum of 3 objects.
		MatchedObject[] objs = FlatMatcher.Find(binarySrc, binaryTempl, count: 3, templScore: 0.9, shapeScore: 0.1);*/

		Image<Bgr, byte> result = src.ToImage<Bgr, byte>();

		if (obj != null)
		{
			// Draw the rotated rectangle.
			System.Drawing.Point[] points = obj.RotatedRect.GetVertices().Select(pF => System.Drawing.Point.Round(pF)).ToArray();
			CvInvoke.Polylines(result, points, true, new MCvScalar(0, 255, 0), 1, LineType.AntiAlias);

			// Draw the center point.
			System.Drawing.Point center = System.Drawing.Point.Round(obj.Center);
			CvInvoke.Circle(result, center, 1, new MCvScalar(0, 255, 0), 1, LineType.AntiAlias);

			// Draw the angle text (the angle difference from the template).
			CvInvoke.PutText(result, obj.Angle.ToString("F1"), center, FontFace.HersheySimplex, 1, new MCvScalar(0,0,255), 1, LineType.AntiAlias);
		}

		CvInvoke.Imshow("result", result);
	}
}

License

This project is licensed under the MIT License - see the LICENSE file for details.