OpenCVSharp_在C#中使用OpenCV_[以Opencv的7大追踪算法为例]

本篇博客应该是属于开荒，因为很难找到C#版Opencv的文章。
本文会详细讲解如何一步步配置OPENCVSHARP（C#中的OPENCV），并给出三个demo，分别是追踪算法CamShift以及Tracker在.NET C#中的实现，以及OPENCV 图像类OpenCvSharp.Mat与C# 图像类System.Drawing.Bitmap的互相转换。
任意新建一个控制台程序，然后打开Nuget包管理器，搜索OpenCvSharp，选择那个头像为猿猴脸的那个库。为什么在众多OpenCvSharp库中选择这个库，是因为其他库我都经过了测试，不能跑。
如下图：

首先声明下，OpenCvSharp是对Opencv C++的封装，以便其能够在.Net中使用。
引入该库 OpenCvSharp；
using OpenCvSharp;

我在这里直接上源代码了，我对一些OPENCV C++ 的程序在C# OpenCvSharp 中进行了重写，下面分别是两个例子：
追踪_OpenCVSharp_Tracker；
Camshift交互式追踪CSharp版；
关于如何使用，两个程序都需要你在OPENCV的窗体上画一个矩形ROI框，移动物体，该方框就会跟随物体而移动，并实时输出该方框左上角坐标。
关于该opencv的图像处理讲解，请大家自己查一下。

//追踪_OpenCVSharp_Tracker；
using System;
using System.Drawing;
using System.Drawing.Imaging;
using OpenCvSharp;
using OpenCvSharp.Tracking;

namespace 追踪_OpenCVSharp_Tracker
{
    class Program
    {
        private static Mat image = new Mat();
        private static OpenCvSharp.Point originPoint = new OpenCvSharp.Point();
        private static Rect2d selectedRect = new Rect2d();
        private static bool selectRegion = false;
        private static int trackingFlag = 0;
        private static void OnMouse(MouseEvent Event, int x, int y, MouseEvent Flags, IntPtr ptr)
        {
            if (selectRegion)
            {
                selectedRect.X = Math.Min(x, originPoint.X);
                selectedRect.Y = Math.Min(y, originPoint.Y);
                selectedRect.Width = Math.Abs(x - originPoint.X);
                selectedRect.Height = Math.Abs(y - originPoint.Y);

                selectedRect = selectedRect & new Rect2d(0, 0, image.Cols, image.Rows);
            }

            switch (Event)

            {
                case MouseEvent.LButtonDown:
                    originPoint = new OpenCvSharp.Point(x, y);
                    selectedRect = new Rect2d(x, y, 0, 0);
                    selectRegion = true;
                    break;

                case MouseEvent.LButtonUp:
                    selectRegion = false;
                    if (selectedRect.Width > 0 && selectedRect.Height > 0)
                    {
                        trackingFlag = -1;
                    }
                    break;
            }
        }
       
        static void Main(string[] args)
        {
            TrackerKCF tracker_KCF= TrackerKCF.Create();
            TrackerMIL trackerMIL = TrackerMIL.Create();
            //cv::Ptr<cv::Tracker> tracker = TrackerCSRT.Create();
            TrackerMedianFlow trackerMedianFlow = TrackerMedianFlow.Create();
            TrackerMOSSE trackerMOSSE = TrackerMOSSE.Create();
            TrackerTLD trackerTLD = TrackerTLD.Create();           
            VideoCapture cap = new VideoCapture();
            cap.Open(0);
            
            if (cap.IsOpened())
            {
                string windowName = "KCF Tracker";
                string windowName2 = "OriginFrame";
                Mat frame = new Mat();
                Mat outputMat = new Mat();
                Cv2.NamedWindow(windowName, 0);
                Cv2.NamedWindow(windowName2, 0);
                Cv2.SetMouseCallback(windowName, OnMouse, new IntPtr());
                while (true)
                {
                    cap.Read(frame);

                    // Check if 'frame' is empty
                    if (frame.Empty())
                    {
                        break;
                    }
                    frame.CopyTo(image);
                    if (trackingFlag != 0)
                    {
                        tracker_KCF.Init(frame, selectedRect);
                        tracker_KCF.Update(frame, ref  selectedRect);
                        frame.CopyTo(outputMat);
                        Rect rect = new Rect((int)selectedRect.X, (int)selectedRect.Y, (int)selectedRect.Width, (int)selectedRect.Height);
                        Console.WriteLine(rect.X+"  "+rect.Y);
                        Cv2.Rectangle(outputMat, rect, new Scalar(255, 255, 0), 2);
                        Cv2.ImShow(windowName2, outputMat);
                    }
                    if (selectRegion && selectedRect.Width > 0 && selectedRect.Height > 0)
                    {
                        Mat roi = new Mat(image, new Rect((int)selectedRect.X, (int)selectedRect.Y, (int)selectedRect.Width, (int)selectedRect.Height));
                        Cv2.BitwiseNot(roi, roi);
                    }
                    Cv2.ImShow(windowName, image);
                    int ch = Cv2.WaitKey(25);
                    if (ch == 27)
                    {
                        break;
                    }
                }
            }
         }
    }
}

//Camshift交互式追踪CSharp版；
using OpenCvSharp;
using System;

namespace 交互式追踪CSharp版
{
    class Program

    {
        private static Mat image = new Mat();
        private static Point originPoint = new Point();
        private static Rect selectedRect = new Rect();
        private static bool selectRegion = false;
        private static int trackingFlag = 0;
        //private static CvMouseCallback callBackFunc = new CvMouseCallback(OnMouse);
        private static void OnMouse(MouseEvent Event, int x, int y, MouseEvent Flags, IntPtr ptr)
        {
            if (selectRegion)
            {
                selectedRect.X = Math.Min(x, originPoint.X);
                selectedRect.Y = Math.Min(y, originPoint.Y);
                selectedRect.Width = Math.Abs(x - originPoint.X);
                selectedRect.Height = Math.Abs(y - originPoint.Y);

                selectedRect = selectedRect & new Rect(0, 0, image.Cols, image.Rows);
            }

            switch (Event)

            {
                case MouseEvent.LButtonDown:
                    originPoint = new Point(x, y);
                    selectedRect = new Rect(x, y, 0, 0);
                    selectRegion = true;
                    break;

                case MouseEvent.LButtonUp:
                    selectRegion = false;
                    if (selectedRect.Width > 0 && selectedRect.Height > 0)
                    {
                        trackingFlag = -1;
                    }
                    break;
            }
        }
        static void Main(string[] args)
        {
            VideoCapture cap = new VideoCapture();
            cap.Open(0);
            if (cap.IsOpened())
            {
                int ch;
                Rect trackingRect = new Rect();

                // range of values for the 'H' channel in HSV ('H' stands for Hue)
                Rangef hist_range = new Rangef(0.0f, 180.0f);
                Rangef[] histRanges = { hist_range };
                //const float* histRanges = hueRanges;

                // min value for the 'S' channel in HSV ('S' stands for Saturation)
                int minSaturation = 40;

                // min and max values for the 'V' channel in HSV ('V' stands for Value)
                int minValue = 20, maxValue = 245;

                // size of the histogram bin
                int[] histSize = { 8 };

                string windowName = "CAMShift Tracker";
                //string windowNameTest = "Test";
                Cv2.NamedWindow(windowName, 0);
                //Cv2.NamedWindow(windowNameTest, 0);
                Cv2.SetMouseCallback(windowName, OnMouse, new IntPtr());
                Mat frame = new Mat();
                Mat hsvImage = new Mat();
                Mat hueImage = new Mat();
                Mat mask = new Mat();
                Mat hist = new Mat();
                Mat backproj = new Mat();

                // Image size scaling factor for the input frames from the webcam
                double scalingFactor = 1;

                // Iterate until the user presses the Esc key
                while (true)
                {
                    // Capture the current frame
                    cap.Read(frame);

                    // Check if 'frame' is empty
                    if (frame.Empty())
                        break;

                    // Resize the frame
                    Cv2.Resize(frame, frame, new Size(), scalingFactor, scalingFactor, InterpolationFlags.Area);
                    frame.CopyTo(image);

                    // Convert to HSV colorspace
                    Cv2.CvtColor(image, hsvImage, ColorConversionCodes.BGR2HSV);
                    if (trackingFlag != 0)
                    {
                        // Check for all the values in 'hsvimage' that are within the specified range
                        // and put the result in 'mask'
                        Cv2.InRange(hsvImage, new Scalar(0, minSaturation, minValue), new Scalar(180, 256, maxValue), mask);
                        /* # 通俗的来讲，这个函数就是判断hsv中每一个像素是否在[lowerb，upperb]之间，注意集合的开闭。
                # 结果是，那么在mask相应像素位置填上255，反之则是0。即重点突出该颜色
                # 即检查数组元素是否在另外两个数组元素值之间。这里的数组通常也就是矩阵Mat或向量。
                # 要特别注意的是：该函数输出的mask是一幅二值化之后的图像。*/
                        //imshow(windowNameTest, mask);
                        //waitKey(0);
                        // Mix the specified channels
                        int[] channels = { 0, 0 };
                        //cout << hsvImage.depth() << endl;
                        hueImage.Create(hsvImage.Size(), hsvImage.Depth());
                        //cout << hueImage.channels() << endl; ;
                        hueImage = hsvImage.ExtractChannel(0);
                        //Cv2.MixChannels(hsvImage, hueImage, channels);
                        /*mixChannels mixChannels()函数用于将输入数组的指定通道复制到输出数组的指定通道。
                        void mixChannels(
                        const Mat* src, //输入数组或向量矩阵,所有矩阵的大小和深度必须相同。
                        size_t nsrcs, //矩阵的数量
                        Mat* dst, //输出数组或矩阵向量,大小和深度必须与src[0]相同
                        size_t ndsts,//矩阵的数量
                        const int* fromTo,//指定被复制通道与要复制到的位置组成的索引对
                        size_t npairs //fromTo中索引对的数目*/
                        if (trackingFlag < 0)
                        {
                            // Create images based on selected regions of interest

                            Mat roi = new Mat(hueImage, selectedRect);
                            Mat maskroi = new Mat(mask, selectedRect);
                            Mat[] roi_source = { roi };
                            int[] channels_ = { 0 };
                            // Compute the histogram and normalize it
                            Cv2.CalcHist(roi_source, channels_, maskroi, hist, 1, histSize, histRanges);
                            Cv2.Normalize(hist, hist, 0, 255, NormTypes.MinMax);

                            trackingRect = selectedRect;
                            trackingFlag = 1;
                        }
                        Mat[] hueImgs = { hueImage };
                        int[] channels_back = { 0 };
                        // Compute the histogram back projection
                        Cv2.CalcBackProject(hueImgs, channels_back, hist, backproj, histRanges);
                        backproj &= mask;
                        //TermCriteria criteria = new TermCriteria(CriteriaTypes.Eps | CriteriaTypes.MaxIter, 10, 1);
                        RotatedRect rotatedTrackingRect = Cv2.CamShift(backproj, ref trackingRect, new TermCriteria(CriteriaType.Eps | CriteriaType.MaxIter, 10, 1));

                        // Check if the area of trackingRect is too small
                        if ((trackingRect.Width * trackingRect.Height) <= 1)
                        {
                            // Use an offset value to make sure the trackingRect has a minimum size
                            int cols = backproj.Cols, rows = backproj.Rows;
                            int offset = Math.Min(rows, cols) + 1;
                            trackingRect = new Rect(trackingRect.X - offset, trackingRect.Y - offset, trackingRect.X + offset, trackingRect.Y + offset) & new Rect(0, 0, cols, rows);
                        }

                        // Draw the ellipse on top of the image
                        Cv2.Ellipse(image, rotatedTrackingRect, new Scalar(0, 255, 0), 3, LineTypes.Link8);
                    }


                    // Apply the 'negative' effect on the selected region of interest
                    if (selectRegion && selectedRect.Width > 0 && selectedRect.Height > 0)
                    {
                        Mat roi = new Mat(image, selectedRect);
                        Cv2.BitwiseNot(roi, roi);
                    }
                    // Display the output image
                    Cv2.ImShow(windowName, image);

                    // Get the keyboard input and check if it's 'Esc'
                    // 27 -> ASCII value of 'Esc' key

                    ch = Cv2.WaitKey(25);
                    if (ch == 27)
                    {
                        break;
                    }
                }
            }
        }
    }
}

另外i，还有一个.Net FrameWork Winform 窗体实时演示摄像机的画面，并包括OpenCVSharp.Mat 类与System.Drawing.Bitmap类的互相转换。
该Demo的窗体界面：

该demo代码：

using System;
using System.Collections.Generic;
using System.ComponentModel;
using System.Data;
using System.Drawing;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using System.Windows.Forms;
using OpenCvSharp;
using System.Drawing.Imaging;
using System.Runtime.InteropServices;

namespace OpenCVSharp_Mat与Bitmap的转换
{
    public partial class Form1 : Form
    {
        VideoCapture cap;
        Mat frame = new Mat();
        Mat dstMat = new Mat();
        Bitmap bmp;
        public Form1()
        {
            InitializeComponent();
        }
        public static Bitmap MatToBitmap(Mat dst)

        {
            return new Bitmap(dst.Cols, dst.Rows, (int)dst.Step(), PixelFormat.Format24bppRgb, dst.Data);

        }
        public static Mat BitmapToMat(Bitmap srcbit)
        {
            int iwidth = srcbit.Width;
            int iheight = srcbit.Height;
            int iByte = iwidth * iheight * 3;
            byte[] result = new byte[iByte];
            int step;
            Rectangle rect = new Rectangle(0, 0, iwidth, iheight);
            BitmapData bmpData = srcbit.LockBits(rect, ImageLockMode.ReadWrite, srcbit.PixelFormat);
            IntPtr iPtr = bmpData.Scan0;
            Marshal.Copy(iPtr, result, 0, iByte);
            step = bmpData.Stride;
            srcbit.UnlockBits(bmpData);
            return new Mat(srcbit.Height, srcbit.Width, new MatType(MatType.CV_8UC3), result, step);
        }
        private void btnRun_Click(object sender, EventArgs e)
        {
            timer1.Enabled = true;
        }

        private void timer1_Tick(object sender, EventArgs e)
        {
            if (cap.IsOpened())
            {
                cap.Read(frame);
                bmp = MatToBitmap(frame);
                pictureBox1.Image = bmp;
                dstMat = BitmapToMat(bmp);
                Cv2.ImShow("dstMat", dstMat);
                //Cv2.WaitKey();
           }
        }

        private void Form1_Load(object sender, EventArgs e)
        {
            try
            {
                cap = new VideoCapture();
                cap.Open(0);
            }
            catch (Exception ex)
            {
                MessageBox.Show(ex.Message);
            }
           
        }

        private void Form1_FormClosed(object sender, FormClosedEventArgs e)
        {
            timer1.Enabled = false;
            if (cap.IsOpened())
            {
                cap.Dispose();
            }
        }
    }
}

里面涉及到一些数字图像处理的函数，大家不懂得可以自己去翻书或者上网查阅资料。多多单步运行，你就会发现他为什么要这样写。具体理论知识篇幅较长，一晚上都讲不完，请大家自行查资料。

如果大家能看到这里，相比是非常喜欢这篇博客了，也对UP主很认可。

那就请关注点赞加收藏吧。
本文图像类的转换部分参考以下博客：
https://blog.csdn.net/qq_34455723/article/details/90053593