OpenCVSharp 4.5 模板匹配

用OpenCVSharp 4.5跑一遍 OpenCV 官方教程

原官方教程链接：OpenCV: Template Matching

核心函数：matchTemplate()， minMaxLoc()

using OpenCvSharp;
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;

namespace ConsoleApp1
{
    class tutorial22 : ITutorial
    {
        static Mat img = new Mat();
        static Mat templ = new Mat();
        static Mat result = new Mat();
        static Mat mask = new Mat();

        const string image_window = "Source Image";
        const string result_window = "Result window";        
        static int match_method;
        static bool use_mask;
        static int max_Trackbar = 5;


        public void Run()
        {
            img = Cv2.ImRead("I:\\csharp\\images\\lena.png",ImreadModes.Color);
            templ = Cv2.ImRead("I:\\csharp\\images\\template_lena.png",ImreadModes.Color);
            mask = Cv2.ImRead("I:\\csharp\\images\\mask_lena.png",ImreadModes.Grayscale);
            if (img.Empty() || templ.Empty())
                return;
            if(!mask.Empty())   
            {
                use_mask = true;
                mask.ConvertTo(mask, MatType.CV_8UC1);
            }

            Cv2.NamedWindow(image_window, WindowFlags.AutoSize);
            Cv2.NamedWindow(result_window, WindowFlags.AutoSize);
            Cv2.NamedWindow("Template", WindowFlags.AutoSize);
            Cv2.NamedWindow("Mask", WindowFlags.AutoSize);
            const string trackbar_label = "Method: \r\n 0: SQDIFF \r\n 1: SQDIFF NORMED \r\n 2: TM CCORR \r\n 3: TM CCORR NORMED \r\n 4: TM COEFF \r\n 5: TM COEFF NORMED";
            Cv2.CreateTrackbar(trackbar_label, image_window, ref match_method, max_Trackbar, MatchingMethod);
            MatchingMethod(0, IntPtr.Zero);
            Cv2.WaitKey(0);

        }

        static void MatchingMethod(int pos, IntPtr userData)
        {
            Mat img_display = new Mat();
            img.CopyTo(img_display);
            int result_cols = img.Cols - templ.Cols + 1;
            int result_rows = img.Rows - templ.Rows + 1;
            result.Create(result_rows, result_cols, MatType.CV_32FC1);
            bool method_accepts_mask = ( TemplateMatchModes.SqDiff == (TemplateMatchModes)match_method || (TemplateMatchModes)match_method == TemplateMatchModes.CCorrNormed);

            if (use_mask && method_accepts_mask)
            { 
                Cv2.MatchTemplate(img, templ, result, (TemplateMatchModes)match_method, mask); 
            }
            else
            { 
                Cv2.MatchTemplate(img, templ, result, (TemplateMatchModes)match_method); 
            }

            Cv2.Normalize(result, result, 0, 1, NormTypes.MinMax, -1, null);
            double minVal; double maxVal; Point minLoc; Point maxLoc;
            Point matchLoc;
            Cv2.MinMaxLoc(result, out minVal, out maxVal, out minLoc, out maxLoc, null);

            if ((TemplateMatchModes)match_method == TemplateMatchModes.SqDiff || (TemplateMatchModes)match_method == TemplateMatchModes.SqDiffNormed)
            { matchLoc = minLoc; }
            else
            { matchLoc = maxLoc; }
            Scalar color = new Scalar(0, 0, 0);
            Cv2.Rectangle(img_display, matchLoc, new Point(matchLoc.X + templ.Cols, matchLoc.Y + templ.Rows), new Scalar(0, 0, 0), 2, LineTypes.Link8, 0);
            Cv2.Rectangle(result, matchLoc, new Point(matchLoc.X + templ.Cols, matchLoc.Y + templ.Rows), new Scalar(0, 0, 0), 2, LineTypes.Link8, 0);
            Cv2.ImShow(image_window, img_display);
            Cv2.ImShow(result_window, result);
            Cv2.ImShow("Template", templ);
            Cv2.ImShow("Mask", mask);
        }
    }
}