【无标题】

@TOCOpenCV实现边缘模板匹配算法
背景概述

OpenCV中自带的模板匹配算法，完全是像素基本的模板匹配，特别容易受到光照影响，光照稍微有所不同，该方法就会歇菜了！搞得很多OpenCV初学者刚学习到该方法时候很开心，一用该方法马上很伤心，悲喜交加，充分感受到了理想与现实的距离，不过没关系，这里介绍一种新的模板匹配算法，主要是基于图像边缘梯度，它对图像光照与像素迁移都有很强的抗干扰能力，据说Halcon的模板匹配就是基于此的加速版本，在工业应用场景中已经得到广泛使用。
结论：大图速度慢，效果不好，小图可以用一下
`在这 Mat edge;
//GaussianBlur(templateMat, bw, Size(7,7), 0, 0);
Mat bw;
threshold(templateMat, bw, 20, 255, CV_THRESH_OTSU);
Canny(bw, edge, 25, 190);
/namedWindow(“Canny”,0);
resizeWindow(“Canny”, 1540, 1180);
imshow(“Canny”, edge);
waitKey(0);/
vector contours;
vector hierarchy;
findContours(edge, contours, CV_RETR_TREE, CV_CHAIN_APPROX_NONE); //找轮廓

Mat gx, gy;
Sobel(fardist3, gx, CV_32F, 1, 0);
Sobel(fardist3, gy, CV_32F, 0, 1);
Mat magnitude, direction;
cartToPolar(gx, gy, magnitude, direction);
long contoursLength = 0;
double magnitudeTemp = 0;
int originx = contours[0][0].x;
int originy = contours[0][0].y;

typedef struct my
{
    int DerivativeX;
    int DerivativeY;
    double Magnitude;
    double MagnitudeN;
}ptin;
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

// 提取dx\dy\mag\log信息
vector contoursInfo;
// 提取相对坐标位置
vector contoursRelative;

// 开始提取
for (int i = 0; i < contours.size(); i++) {
    int n = contours[i].size();
    contoursLength += n;
    contoursInfo.push_back(vector(n));
    vector points(n);
    for (int j = 0; j < n; j++) {
        int x = contours[i][j].x;
        int y = contours[i][j].y;
        points[j].x = x - originx;
        points[j].y = y - originy;
        ptin pointInfo;
        pointInfo.DerivativeX = gx.at(y, x);
        pointInfo.DerivativeY = gy.at(y, x);
        magnitudeTemp = magnitude.at(y, x);
        pointInfo.Magnitude = magnitudeTemp;
        if (magnitudeTemp != 0)
            pointInfo.MagnitudeN = 1 / magnitudeTemp;
        contoursInfo[i][j] = pointInfo;
    }
    contoursRelative.push_back(points);
}
// 计算目标图像梯度
Mat gradx, grady;
Sobel(fardist3, gradx, CV_32F, 1, 0);
Sobel(fardist3, grady, CV_32F, 0, 1);
Mat mag, angle;
cartToPolar(gradx, grady, mag, angle);
// NCC模板匹配
double minScore = 0.99;		//deafult value
double greediness = 0.8;		//deafult value
double nGreediness = 0.99;		//deafult value
double nMinScore = 0.99;		//deafult value

double partialScore = 0;
double resultScore = 0;
int resultX = 0;
int resultY = 0;
double start = (double)getTickCount();
for (int row = 0; row < fardist3.rows; row++)
{
    for (int col = 0; col < fardist3.cols; col++) {
        double sum = 0;
        long num = 0;
        for (int m = 0; m < contoursRelative.size(); m++) {
            for (int n = 0; n < contoursRelative[m].size(); n++) {
                num += 1;
                int curX = col + contoursRelative[m][n].x;
                int curY = row + contoursRelative[m][n].y;
                if (curX < 0 || curY < 0 || curX > fardist3.cols - 1 || curY > fardist3.rows - 1) {
                    continue;
                }

                // 目标边缘梯度
                double sdx = gradx.at(curY, curX);
                double sdy = grady.at(curY, curX);

                // 模板边缘梯度
                double tdx = contoursInfo[m][n].DerivativeX;
                double tdy = contoursInfo[m][n].DerivativeY;

                // 计算匹配
                if ((sdy != 0 || sdx != 0) && (tdx != 0 || tdy != 0))
                {
                    double nMagnitude = mag.at(curY, curX);
                    if (nMagnitude != 0)
                        sum += (sdx * tdx + sdy * tdy) * contoursInfo[m][n].MagnitudeN / nMagnitude;
                }

                // 任意节点score之和必须大于最小阈值
                partialScore = sum / num;
                if (partialScore < min((minScore - 1) + (nGreediness * num), nMinScore * num))
                    break;
            }
        }

        // 保存匹配起始点
        if (partialScore > resultScore)
        {
            resultScore = partialScore;
            resultX = col;
            resultY = row;
        }
    }
}
cout << resultScore << endl;
cout << resultY << endl;
cout << resultX << endl;
CvPoint point;
point.x = resultX;
point.y = resultY;
circle(fardist3, point, 10, Scalar(255, 255, 255), 8);
namedWindow("src",0);
resizeWindow("src", 1540, 1180);
imshow("src", fardist3);
waitKey(0);里插入代码片`
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