1. 程式人生 > >OpenCV3入門(九)影象幾何變換

OpenCV3入門(九)影象幾何變換

1、影象縮放

假設影象x軸的縮放因子Sx, y軸方向的縮放因子Sy,相應的變換表示式為:

 

函式原型為:

CV_EXPORTS_W void resize( InputArray src, OutputArray dst,
                          Size dsize, double fx = 0, double fy = 0,
                          int interpolation = INTER_LINEAR );

示例程式如下。

img = imread("D:\\WORK\\5.OpenCV\\LeanOpenCV\\pic_src\\pic9.bmp");
imshow("原圖", img);
resize(img, img2, Size(), 0.5, 0.5);
imshow("縮放圖1", img2);
resize(img, img3, Size(), 0.8, 0.5);
imshow("縮放圖2", img3);

執行效果如下圖。

 

 resize(img, img2, Size(), 1.2, 1.2);

 

2、影象平移

假設影象x軸的平移量Tx, y軸方向的平移量Ty,相應的變換表示式為:

 

仿射變換的原理為:

dst(x,y)=src(M11x+M12y+M13,M21x+M22y+M23)

平移操作可以使用OpenCV的仿射變換函式來實現,使用的變換矩陣為:

 

函式原型為:

CV_EXPORTS_W void warpAffine( InputArray src, OutputArray dst,
                              InputArray M, Size dsize,
                              int flags = INTER_LINEAR,
                              int borderMode = BORDER_CONSTANT,
                              const Scalar& borderValue = Scalar());

影象平移示例。

img = imread("D:\\WORK\\5.OpenCV\\LeanOpenCV\\pic_src\\pic9.bmp", IMREAD_GRAYSCALE);
imshow("原圖", img);

// x軸平移20,y軸平移10, 2 * 3矩陣
Mat M = Mat::zeros(2, 3, CV_32FC1);
M.at<float>(0, 0) = 1;
M.at<float>(0, 2) = 20;
M.at<float>(1, 1) = 1;
M.at<float>(1, 2) = 10;
warpAffine(img, img2, M, img.size());
imshow("平移圖", img2);

3、影象旋轉

假設點P0(x0,y0),角度為a,令L=|OP|=sqrt(x*x + y*y).

P0選擇b度到P1(x1,y1),則

x1=L*cos(a+b)=L* cos(a)*cos(b) – L*sin(a)*sin(b) = x0*cos(b) - y0*sin(b)

y1=L*sin(a+b)=L* sin(a)*cos(b) + L*cos(a)*sin(b) = y0*cos(b) + x0*sin(b)

OpenCV內建仿射變換的旋轉函式,支援任意點為中心的影象旋轉,函式原型為:

CV_EXPORTS_W Mat getRotationMatrix2D( Point2f center, double angle, double scale );

示例程式碼如下。

img = imread("D:\\WORK\\5.OpenCV\\LeanOpenCV\\pic_src\\pic9.bmp");
imshow("原圖", img);

Point center = Point(img.cols / 2, img.rows / 2);
Mat m1= getRotationMatrix2D(center, 30, 1.0);
Mat m2 = getRotationMatrix2D(center, 30, 0.7);
Mat m3 = getRotationMatrix2D(center, 30, 1.2);
warpAffine(img, img1, m1, img.size());
warpAffine(img, img2, m2, img.size());
warpAffine(img, img3, m3, img.size());
imshow("img1", img1);
imshow("img2", img2);
imshow("img3", img3);

修改旋轉角度效果如下圖。

Mat m1= getRotationMatrix2D(center, 180, 1.0);

Mat m2 = getRotationMatrix2D(center, 270, 0.7);

 

如果旋轉點的座標原點不在圖片中心,則圖片繞著指定點旋轉。

Point center = Point(0, 0);

Mat m1= getRotationMatrix2D(center, 30, 1.0);

Mat m2 = getRotationMatrix2D(center,-45, 1.0);

對應的矩陣為:

m1=

[0.8660254037844387, 0.4999999999999999, 0;

 -0.4999999999999999, 0.8660254037844387, 0]

m2=

[0.7071067811865476, -0.7071067811865476, 0;

 0.7071067811865476, 0.7071067811865476, 0]

輸出效果如下圖。

4、影象重對映

重對映就是把一個影象中一個為之的畫素放置到另一個圖片指定位置過程。由於對映後的影象在原圖中可能沒有對應的整數座標點畫素,所以為了完成重對映需要做一些插值作為非整數畫素座標。我們通過重對映來表達每個畫素的位置(x, y):g(x, y)=f(h(x,y))

OpenCV使用remap函式完成重對映功能,函式原型為:

CV_EXPORTS_W void remap( InputArray src, OutputArray dst,
                         InputArray map1, InputArray map2,
                         int interpolation, int borderMode = BORDER_CONSTANT,
                         const Scalar& borderValue = Scalar());

測試程式碼如下。

void update_map(int ind, Mat &map_x, Mat &map_y)
{
    for (int i = 0; i < map_x.rows; i++)
    {
        for (int j = 0; j < map_x.cols; j++)
        {
            switch (ind)
            {
            case 0:
                if (j > map_x.cols*0.25 && j < map_x.cols*0.75 && i > map_x.rows*0.25 && i < map_x.rows*0.75)
                {
                    map_x.at<float>(i, j) = 2 * (j - map_x.cols*0.25f) + 0.5f;
                    map_y.at<float>(i, j) = 2 * (i - map_x.rows*0.25f) + 0.5f;
                }
                else
                {
                    map_x.at<float>(i, j) = 0;
                    map_y.at<float>(i, j) = 0;
                }
                break;
            case 1:
                map_x.at<float>(i, j) = (float)j;
                map_y.at<float>(i, j) = (float)(map_x.rows - i);
                break;
            case 2:
                map_x.at<float>(i, j) = (float)(map_x.cols - j);
                map_y.at<float>(i, j) = (float)i;
                break;
            case 3:
                map_x.at<float>(i, j) = (float)(map_x.cols - j);
                map_y.at<float>(i, j) = (float)(map_x.rows - i);
                break;
            default:
                break;
            } // end of switch
        }
    }
}

int main() {
    Mat src = imread("D:\\WORK\\5.OpenCV\\LeanOpenCV\\pic_src\\pic9.bmp");
    imshow("原圖", src);

    Mat dst(src.size(), src.type());
    Mat map_x(src.size(), CV_32FC1);
    Mat map_y(src.size(), CV_32FC1);

    update_map(0, map_x, map_y);
    remap(src, img1, map_x, map_y, INTER_LINEAR, BORDER_CONSTANT, Scalar(0, 0, 0));
    imshow("img1", img1);

    update_map(1, map_x, map_y);
    remap(src, img2, map_x, map_y, INTER_LINEAR, BORDER_CONSTANT, Scalar(0, 0, 0));
    imshow("img2", img2);

    update_map(2, map_x, map_y);
    remap(src, img3, map_x, map_y, INTER_LINEAR, BORDER_CONSTANT, Scalar(0, 0, 0));
    imshow("img3", img3);

    update_map(3, map_x, map_y);
    remap(src, img4, map_x, map_y, INTER_LINEAR, BORDER_CONSTANT, Scalar(0, 0, 0));
    imshow("img4", img4);

    waitKey();
}

輸入影象:

程式碼實現四種remap效果。

 

remap後的影象:

 

例子2:x軸不壓縮,y軸按照一元二次曲線進行壓縮,對稱抽為src.rows / 2。當y= src.rows / 2,對應變換前的圖座標src.rows,所以影象被壓縮。當y=[ src.rows / 2,src.rows]時,y軸被反轉。

for (int i = 0; i < src.rows; i++)
{
    for (int j = 0; j < src.cols; j++)
    {
        map_x.at<float>(i, j) = j;
        map_y.at<float>(i, j) = (float)((-1 * pow(i- src.rows / 2, 2) / pow(src.rows / 2, 2)) + 1) * src.rows;
    }
}

5、參考文獻

1、《OpenCV3 程式設計入門》,電子工業出版社,毛星雨著

2、《學習OpenCV》,清華大學出版社,Gary Bradski, Adrian kaehler著

3、Remapping

https://docs.opencv.org/3.4/d1/da0/tutorial_remap.html

4、OpenCV影象旋轉

https://www.cnblogs.com/konglongdanfo/p/9135501.html

 

技術部落格,轉載請註明。

https://www.cnblogs.com/pingwen/p/12329168.html