opencv 学习笔记（十五） 第三周总结

//1.包含必要的头文件；
#include<opencv.hpp>
#include<iostream>
using namespace std;
using namespace cv;
#if 1
void tong(Mat &img1, Mat &img2, Mat &img3)
{
	int r = img1.rows;
	int c = img1.cols + 5 + img2.cols;
	img3.create(r, c, img1.type());
	img1.copyTo(img3(Rect(0, 0, img1.cols, img1.rows)));
	img2.copyTo(img3(Rect(img1.cols, 0, img2.cols, img2.rows)));
	imshow("同一窗口", img3);
}
int main()
{
	Mat img;
	img = imread("automobile.jpg",0);
	//2.输出图像automobile.jpg的宽度和高度；
	cout << "宽度" << img.cols << endl;
	cout << "高度" << img.rows << endl;
	//3.按照幂律变换公式s = crv，取c = 1.3，v = 2.5对图像automobile.jpg进行灰度变换； 8:39
	Mat m(img.size(), img.type());
	for (int i = 0; i < img.rows; i++)
	{
		for (int j = 0; j < img.cols; j++)
		{
			m.at<uchar>(i, j) = saturate_cast<uchar>(1.3*(pow(img.at<uchar>(i, j), 2.5)));
		}
	}
	imshow("幂律变换", m);
	// 4.利用equalizeHist()图像automobile.jpg进行直方图均衡化，并存储输出结果图像； 8:46 修改9:26
	int bins = 256;
	Mat hist;
	int hist_size[] = { bins };
	float range[] = { 0,256 };
	const float *ranges[] = { range };
	calcHist(&img, 1, 0, Mat(), hist, 1, hist_size, ranges);
	double max_l;
	minMaxLoc(hist, 0, &max_l);
	int sc = 2;
	int gg = 256;
	Mat z = Mat::zeros(gg, sc*bins, img.type());
	for (int i = 0; i < bins; i++)
	{
		float bin_v = hist.at<float>(i);
		int zg = cvRound(bin_v*gg / max_l);
		rectangle(z, Point(i*sc, gg - 1), Point((i + 1)*sc-1 , gg - zg), Scalar(255,255,255));
	}
	imshow("直方图", z);

	Mat j;
	equalizeHist(z, j);
	imshow("直方图均衡化", j);

	//5.对automobile.jpg进行灰度处理，存储为automobile_gray.jpg； 8:48
	Mat img2;
	Mat img1 = imread("automobile.jpg");
	cvtColor(img1, img2, COLOR_BGR2GRAY);
	imshow("灰度图", img2);
	imwrite("automobile_gray.jpg", img2);

	//6.计算灰度图automobile_gray.jpg像素值为90的像素点数，并输出结果； 8:49
	int sum = 0;
	for (int i = 0; i < img2.rows; i++)
	{
		for (int j = 0; j < img2.cols; j++)
		{
			if (img2.at<uchar>(i, j) == 90)
			{
				sum++;
			}
		}
	}
	cout << "像素值为90的像素点数" << sum << endl;
	//7.在图像automobile.jpg四周添加10个像素宽的蓝色边框并显示； 8:51
	Mat bk;
	copyMakeBorder(img1, bk, 10, 10, 10, 10, 0, Scalar(255, 0, 0));
	imshow("边框", bk);
	//8.检测图像automobile.jpg的边缘并显示； 8:57 修改 9:21
	Mat g1;
	GaussianBlur(img, g1, Size(3, 3), 3);
	Canny(g1, g1, 10, 10);
	imshow("边缘检测", g1);
	//9.在单窗口中同时展示图像automobile.jpg和automobile_gray.jpg； 9:00
	Mat d1 = imread("automobile.jpg");
	Mat d2 = imread("automobile_gray.jpg");
	Mat d3;
	tong(d1, d2, d3);
	//10.采用适当函数对automobile.jpg进行图像分割，并显示分割后效果； 9:02
	Mat fg;
	threshold(img, fg, 0, 255, THRESH_OTSU);
	imshow("分割", fg);
	//11.构造宽度为400，高度为300的全0矩阵B；9:03
	Mat B = Mat::zeros(300, 400, d1.type()); 
	//12.在automobile.jpg中截取(50, 200)、(450, 200)、(450, 500)、(50, 500)所围区域，存储为“automobile_part.jpg”的新图像； 9:11
	Rect rect(50,200 , 400, 300);
	cout << rect.tl() << endl;
	Mat x = d1(rect);
	imshow("截取", x);
	imwrite("automobile_part.jpg", x);
	//13.用automobile_part.jpg像素值替换矩阵B的对应像素值；9:13
	B = x.clone();
	imshow("替换后",B);
	//14.计算automobile_part.jpg中心点旋转（旋转角度15°）的旋转矩阵并打印； 9:14
	Point center(B.cols / 2, B.rows / 2);
	Mat xz = getRotationMatrix2D(center, 15, 1);
	cout << xz << endl;
	//15.显示automobile_part.jpg中心点旋转（旋转角度15°）的旋转效果；9:16
	Mat xz1;
	warpAffine(B, xz1, xz, xz1.size());
	imshow("旋转效果", xz1);
	//16.对图像automobile.jpg在x轴方向缩小1 / 2，在y轴方向上都缩小1 / 3，显示缩小后效果；9:17
	Mat xs;
	resize(d1, xs, Size(d1.cols / 2, d1.rows / 3));
	imshow("缩小", xs);
	//17.用适当方法对图像automobile.jpg进行图像分割，显示图像分割效果； 9:18
	Mat fg1;
	threshold(img, fg1, 0, 255, THRESH_OTSU);
	imshow("分割2", fg1);
	18.用高斯滤波器对automobile.jpg进行平滑处理； 9:19
	Mat G;
	GaussianBlur(img, G, Size(7, 7), 7);
	imshow("高斯滤波", G);
	//19.提取automobile.jpg的轮廓并显示； 9:21
	vector<vector<Point>> m1;
	Mat g;
	GaussianBlur(img, g, Size(3, 3), 3);
	Canny(g, g, 10, 10);
	Mat lun = Mat::zeros(img.size(), 0);
	findContours(g, m1, RETR_LIST, CHAIN_APPROX_SIMPLE, Point());
	drawContours(lun, m1, -1, Scalar(255));
	imshow("轮廓", lun);
	waitKey(0);
}
#endif // 0