CSharp + OpenCvSharp实现医学图像拼接_csharp cvzone

基于网络上很多都是C++ 、Python实现的图像拼接，或者直接调用的Stitch函数进行的拼接，想着手动实现一个C#版本，具体思路如下：

实际拼接效果如下：


实现起来不是很难，但还是有许多可以优化的点，这些放在以后把。
下面展示一些 内联代码片。

/// <summary>
        /// 基于SIFT算法特征提取进行拼接
        /// </summary>
        /// <param name="sender"></param>        
        /// <param name="e"></param>
        private void btnSift_Click(object sender, EventArgs e)
        {
            #region 读取图片
            //这一部分相对而言较为简单，读者按自己需求编写就可            
            #endregion
            //从这开始是正式流程
            /*
             *   图像二值化，加快特征点检测速度
             */
            Mat srcImgBgr2Gray = new Mat();
            Mat testImgBgr2Gray = new Mat();
            Cv2.CvtColor(srcImg, srcImgBgr2Gray,ColorConversionCodes.BGRA2GRAY);
            Cv2.CvtColor(testImg, testImgBgr2Gray, ColorConversionCodes.BGR2GRAY);

            #region 创建 SIFT  
            //var MySift = OpenCvSharp.Features2D.SIFT.Create(400);
            var MySift = OpenCvSharp.XFeatures2D.SIFT.Create(2000);

            /*
             *    特征点
             */
            KeyPoint[] srcKeyPoints = MySift.Detect(srcImgBgr2Gray);
            KeyPoint[] testKeyPoints = MySift.Detect(testImgBgr2Gray);

            /*
             *    描述子
             */
            Mat srcDescriptors = new Mat();
            Mat testDescriptors = new Mat();

            /*
             *    计算特征向量
             */
            srcKeyPoints = MySift.Detect(srcImg);
            testKeyPoints = MySift.Detect(testImg);
            MySift.Compute(srcImgBgr2Gray, ref srcKeyPoints, srcDescriptors);
            MySift.Compute(testImgBgr2Gray, ref testKeyPoints, testDescriptors);

            #endregion

            #region 创建FLANN匹配器
            FlannBasedMatcher matcher = new FlannBasedMatcher();
            DMatch[] matches = matcher.Match(srcDescriptors, testDescriptors);

            #endregion

            #region 比率测试
            double max_dist = 0;
            double min_dist = 100;
            for (int i = 0; i < srcDescriptors.Rows; i++)
            {
                double dist = matches[i].Distance;
                if (dist < 0.9 * min_dist)
                {
                    min_dist = dist;
                }
                if (dist > 0.9 * max_dist)
                {
                    max_dist = dist;
                }
            }            
            List<DMatch> good_matches = new List<DMatch>();
            for (int i = 0; i < srcDescriptors.Rows; i++)
            {
                if (matches[i].Distance < 3 * min_dist)
                {
                    good_matches.Add(matches[i]);   
                }
            }
            if (good_matches.Count <= 0)
            {
                MessageBox.Show("未匹配到特征点，无法进行拼接！\n请尝试手动拼接","提醒");
                return;
            }
            /*
             *    绘制匹配点
             */
            Mat img_matches = new Mat();
            Cv2.DrawMatches(srcImg, srcKeyPoints, testImg, testKeyPoints, good_matches, img_matches);
            
            /*
             *   创建模板
             */
            Mat srcImgTemp = new Mat();
            Cv2.CopyMakeBorder(srcImg, srcImgTemp, top, bottom, left, right, BorderTypes.Constant, 0);
            Mat testImgTemp = new Mat();            
            Cv2.CopyMakeBorder(testImg, testImgTemp, top, bottom, left, right, BorderTypes.Constant, 0); 
            #endregion

            #region 透视变换
            List<Point2f> srcImgTransf = new List<Point2f>();
            List<Point2f> testImgTransf = new List<Point2f>();
            for (int i = 0; i < good_matches.Count; i++)
            {
                srcImgTransf.Add(srcKeyPoints[good_matches[i].QueryIdx].Pt);
                testImgTransf.Add(testKeyPoints[good_matches[i].TrainIdx].Pt);
            }
            List<Point2d> srcImgPoints = srcImgTransf.ConvertAll(Point2fToPoint2d);
            List<Point2d> testImgPoints = testImgTransf.ConvertAll(Point2fToPoint2d);

            /*
             *   单应性矩阵
             */            
            Mat testImgWarp = new Mat();
            Mat M = Cv2.FindHomography(testImgPoints, srcImgPoints, HomographyMethods.Ransac, 5.0);
            OpenCvSharp.Size sizeTest = new OpenCvSharp.Size(testImg.Cols, testImg.Rows + 250);
            //CalcCorners(M, testImg);
            /*
             *   透视变换
             */                     
            Cv2.WarpPerspective(testImgTemp, testImgWarp, M, sizeTest, InterpolationFlags.WarpFillOutliers, BorderTypes.Constant, 0);           
            #endregion            

            #region 图像融合

            /*
             *     ①图像融合
             */
            int dstHeight = testImgWarp.Rows;
            int dstWidth = srcImg.Cols;
            Mat dst = new Mat(dstHeight, dstWidth, MatType.CV_8UC3);
            dst.SetTo(0);

            Rect rect = new Rect(0, 0, testImgWarp.Cols, testImgWarp.Rows);
            Mat dstTestImgWarp = new Mat(dst, rect);
            testImgWarp.CopyTo(dstTestImgWarp);

            Rect rectTestImg = new Rect(0, 0, srcImg.Cols, srcImg.Rows);
            Mat dstTestImg = new Mat(dst, rectTestImg);
            srcImg.CopyTo(dstTestImg);

            /*
             *     ②黑边处理
             */
            //OptimizeSeam(srcImg, dstTestImgWarp,dst);
            #endregion

            /*
             *   输出图像
             */            
            Bitmap map = OpenCvSharp.Extensions.BitmapConverter.ToBitmap(dst);
            pbStitchImg.Image = map;
        }
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148

以上便是本次实现过程demo，期待和大家共同探讨。