目录1.环境安装2.示例程序3.认识Mat容器单通道矩阵1. 核心判断依据2. 完整维度信息3. 对比多通道写法区分双通道示例Vec2i2 通道 int三通道示例Vec3b3 通道 uchar4. 补充知识点4.Mat类元素的读取5.Mat支持的运算6.图像加载与显示imread()nameWindow()imshow()imwrite()7.Image watch 使用8.视频加载摄像头使用VideoCapture()get()VideoWriter()9.图像颜色空间变换converTo()cvtColor()10.多通道分离与合并split()merge()11.两张图像像素比较min()max()minMaxLoc()12.图像像素逻辑操作bitwise_not()bitwise_and()bitwise_or()bitwise_xor()13.图像二值化threshold()adaptiveThreshold()1.环境安装参考之前博客2.示例程序路径如下每个代码是单独的main程序可以新建工程调用。注意有的执行需要带参数有的不需要看源码。exampleedge.cpp边缘检测kmeans.cpp•K聚类camshiftdemo.cpp 跟踪。。。3.认识Mat容器#include opencv2/opencv.hpp #include iostream using namespace cv; using namespace std; int main() { system(color F0); // 3行3列 单通道8位无符号矩阵未初始化随机值 Mat a(3, 3, CV_8UC1); // Size(宽,高) 4列4行 单通道矩阵 Mat b(Size(4, 4), CV_8UC1); // 5行5列 双通道统一填充 Scalar(4,5)第三个参数6丢弃 Mat c0(5, 5, CV_8UC2, Scalar(4, 5, 6)); Mat c1(5, 5, CV_8UC2, Scalar(4, 5, 6)); Mat c2(5, 5, CV_8UC2, Scalar(4, 5, 6)); // 1行5列 int 单行向量 Mat d (cv::Mat_int(1, 5) 1, 2, 3, 4, 5); // 以d为对角线生成5阶对角方阵 Mat e Mat::diag(d); // 截取e行2~3列2~3左闭右开Range(2,4) Mat f Mat(e, Range(2, 4), Range(2, 4)); cout c0 endl; cout c0 endl; cout c1 endl; cout c1 endl; cout c2 endl; cout c2 endl; cout d endl; cout d endl; cout e endl; cout e endl; cout f endl; cout f endl; return 0; }单通道矩阵1. 核心判断依据cpp运行Mat a (cv::Mat_int(3, 3) 1, 2, 3, 4, 5, 6, 7, 8, 9);cv::Mat_int代表单通道整型矩阵模板参数只有一个类型int对应通道数C1每个像素仅存储 1 个 int 数值不存在多通道分组。2. 完整维度信息尺寸3 行 × 3 列通道数1数据类型32 位整型对应CV_32SC1矩阵内容plaintext[1, 2, 3; 4, 5, 6; 7, 8, 9]3. 对比多通道写法区分双通道示例Vec2i2 通道 intcpp运行Mat twoCh (Mat_Vec2i(2,2) Vec2i(1,2), Vec2i(3,4), Vec2i(5,6), Vec2i(7,8));三通道示例Vec3b3 通道 ucharcpp运行Mat threeCh (Mat_Vec3b(2,2) Vec3b(1,2,3), Vec3b(4,5,6), Vec3b(7,8,9), Vec3b(10,11,12));4. 补充知识点Mat_TT 为普通基础类型uchar/int/float→单通道T 为Vec2b/Vec3i/Vec4f向量类型 →多通道等价普通构造写法cpp运行Mat a(3,3, CV_32SC1); // CV_32SC1 32bit int, single channel4.Mat类元素的读取#include opencv2/opencv.hpp #include iostream using namespace cv; using namespace std; int main() { system(color F0); // 3行3列int矩阵流式赋值 Mat a (cv::Mat_int(3, 3) 1, 2, 3, 4, 5, 6, 7, 8, 9); // 单通道 8UC1Scalar仅取第一个值4 Mat c0(5, 5, CV_8UC1, Scalar(4, 5, 6)); // 双通道 8UC2Scalar取前两个值4、5 Mat c1(5, 5, CV_8UC2, Scalar(4, 5, 6)); // 三通道 8UC3Scalar取全部三个值4、5、6 Mat c2(5, 5, CV_8UC3, Scalar(4, 5, 6)); cout a endl; // 读取矩阵a第0行第0列int像素 cout a.atint(0, 0) endl; cout c1 endl; // 读取c1双通道像素(0行1列)存入Vec2b向量 Vec2b vc c1.atVec2b(0, 1); cout vc endl; // 打印通道0数值 cout (int)vc[0] endl; cout c2 endl; // 原始截图后半段截断完整寻址写法取c2第2行第3列像素第0通道 cout (double)(*(c2.data c2.step[0] * 2 c2.step[1] * 2 3)) endl; // 打印Vec2b第二通道 cout (int)vc[1] endl; cout 注意vc[1] 注意 endl; return 0; }5.Mat支持的运算OpenCV: OpenCV modules#include opencv2/opencv.hpp #include iostream using namespace cv; using namespace std; int main() { system(color F0); Mat a (cv::Mat_int(3, 3) 1, 2, 3, 4, 5, 6, 7, 8, 9); Mat b (cv::Mat_int(3, 3) 1, 2, 3, 7, 8, 9, 4, 5, 6); Mat c (cv::Mat_double(3, 3) 1.0, 2.1, 3.2, 4.0, 5.1, 6.2, 2, 2, 2); Mat d (cv::Mat_double(3, 3) 1.0, 2.1, 3.2, 4.0, 5.1, 6.2, 2, 2, 2); cout 两个矩阵的和 endl a b endl; cout 两个矩阵的差 endl c - d endl; cout 矩阵数乘 endl 2 * a endl; cout 矩阵数除 endl d / 2.0 endl; cout 矩阵减数 endl a - 1 endl; cout 两矩阵相乘 endl c*d endl; cout 矩阵内积 endl a.dot(b) endl; cout 矩阵对应位相乘 endl a.mul(b) endl; cout 两个矩阵最小值 endl min(a, b) endl; return 0; }6.图像加载与显示imread()nameWindow()imshow()imwrite()#include opencv2/opencv.hpp #include iostream using namespace cv; using namespace std; int main() { Mat img; img imread(C:/opencv/lena.png,IMREAD_COLOR); Mat gray imread(C:/opencv/lena.png, IMREAD_GRAYSCALE); namedWindow(img, WINDOW_AUTOSIZE); namedWindow(gray, WINDOW_NORMAL); imshow(img, img); imshow(gray, gray); //vectorint compression_params; //compression_params.push_back(IMWRITE_PNG_COMPRESSION); //compression_params.push_back(9); imwrite(C:/opencv/newlena.jpg, img); waitKey(0); return 0; }7.Image watch 使用8.视频加载摄像头使用VideoCapture()get()VideoWriter()#include opencv2/opencv.hpp #include iostream using namespace cv; using namespace std; int main() { /*VideoCapture video; video.open(1); if (!video.isOpened()) { cout 请检查视频文件 endl; return -1; } cout 视频帧率 video.get(CAP_PROP_FPS) endl; cout 视频宽度 video.get(CAP_PROP_FRAME_WIDTH) endl; while (1) { Mat frame; video frame; if (frame.empty()) { break; } imshow(video, frame); uchar c waitKey(1000 / video.get(CAP_PROP_FPS)); if (cq) { break; } }*/ Mat img; VideoCapture video; //使用某个摄像头 video.open(C:/opencv/55.mp4); //VideoCapture video(1); //使用某个摄像头 //读取视频 //VideoCapture video; //video.open(cup.mp4); if (!video.isOpened()) // 判断是否调用成功 { cout 打开摄像头失败请确实摄像头是否安装成功; return -1; } video img; //获取图像 //检测是否成功获取图像 if (img.empty()) //判断有没有读取图像成功 { cout 没有获取到图像 endl; return -1; } bool isColor (img.type() CV_8UC3); //判断相机视频类型是否为彩色 VideoWriter writer; int codec VideoWriter::fourcc(M, J, P, G); // 选择编码格式 double fps 25.0; //设置视频帧率 string filename live.avi; //保存的视频文件名称 writer.open(filename, codec, fps, img.size(), isColor); //创建保存视频文件的视频流 if (!writer.isOpened()) //判断视频流是否创建成功 { cout 打开视频文件失败请确实是否为合法输入 endl; return -1; } while (1) { //检测是否执行完毕 if (!video.read(img)) //判断能都继续从摄像头或者视频文件中读出一帧图像 { cout 摄像头断开连接或者视频读取完成 endl; break; } writer.write(img); //把图像写入视频流 //writer img; imshow(Live, img); //显示图像 char c waitKey(50); if (c 27) //按ESC案件退出视频保存 { break; } } // 退出程序时刻自动关闭视频流 //video.release(); //writer.release(); return 0; }9.图像颜色空间变换converTo()cvtColor()#include opencv2/opencv.hpp #include iostream using namespace cv; using namespace std; int main() { // 读取图片默认BGR三通道CV_8UC3 Mat img imread(C:/opencv/lena.png); Mat img32; // 8UC3 → 32位浮点像素值归一化到 [0, 1] img.convertTo(img32, CV_32F, 1 / 255.0, 0); Mat HSV, HSV32; // 原图8位BGR转HSV cvtColor(img, HSV, COLOR_BGR2HSV); // 浮点归一化BGR转HSV cvtColor(img32, HSV32, COLOR_BGR2HSV); Mat gray0, gray1; // BGR图转灰度正确 cvtColor(img, gray0, COLOR_BGR2GRAY); // 错误img是BGR存储不能用RGB2GRAY灰度结果会失真 cvtColor(img, gray1, COLOR_RGB2GRAY); return 0; }10.多通道分离与合并split()merge()#include opencv2/opencv.hpp #include iostream using namespace cv; using namespace std; int main() { Mat img imread(C:/opencv/lena.png); Mat img32; img.convertTo(img32, CV_32F, 1 / 255.0, 0); Mat gray0, gray1; cvtColor(img, gray0, COLOR_BGR2GRAY); cvtColor(img, gray1, COLOR_RGB2GRAY); Mat imgs[3]; split(img, imgs); Mat img0, img1, img2; img0 imgs[0]; img1 imgs[1]; img2 imgs[2]; Mat img_H; merge(imgs, 3, img_H); Mat zero Mat::zeros(Size(img.cols, img.rows), CV_8UC1); vectorMat imgsV; imgsV.push_back(img0); imgsV.push_back(zero); imgsV.push_back(zero); /*imgsV.push_back(zero); imgsV.push_back(zero); imgsV.push_back(img2);*/ Mat imgsVH; merge(imgsV, imgsVH); return 0; }11.两张图像像素比较min()max()minMaxLoc()#include opencv2/opencv.hpp #include iostream using namespace cv; using namespace std; int main() { Mat img imread(C:/opencv/learnOpenCV/lena.png); Mat white imread(C:/opencv/learnOpenCV/mark_white.png); Mat black imread(C:/opencv/learnOpenCV/mark_black.png); Mat Min, Max; min(img, white, Min); max(img, black, Max); Mat gray, gray_white; cvtColor(img, gray, COLOR_BGR2GRAY); cvtColor(white, gray_white, COLOR_BGR2GRAY); double minVal, maxVal; Point minLoc, maxLoc; minMaxLoc(gray, minVal, maxVal, minLoc, maxLoc, gray_white); return 0; }12.图像像素逻辑操作bitwise_not()bitwise_and()bitwise_or()bitwise_xor()#include opencv2/opencv.hpp #include iostream using namespace cv; using namespace std; int main() { Mat a (Mat_uchar(1, 2) 0, 5); Mat b (Mat_uchar(1, 2) 0, 6); Mat Xor, Or, Not, And; bitwise_not(a, Not); cout a Not Not endl; bitwise_and(a, b, And); bitwise_or(a, b, Or); bitwise_xor(a, b, Xor); cout And endl; cout Or endl; cout Xor endl; Mat img imread(C:/opencv/learnOpenCV/lena.png); Mat mark imread(C:/opencv/learnOpenCV/mark_white.png); Mat result; bitwise_and(img, mark, result); Mat img_inv; //bitwise_not(img, img_inv); cvtColor(mark, mark, COLOR_BGR2GRAY); bitwise_not(img, img_inv, mark); Mat mark_black imread(C:/opencv/learnOpenCV/mark_black.png); min(img, mark_black, img); img img img_inv; return 0; }13.图像二值化threshold()adaptiveThreshold()#include opencv2\opencv.hpp #include iostream using namespace std; using namespace cv; int main() { Mat img imread(C:/opencv/learnOpenCV/lena.png); if (img.empty()) { cout 请确认图像文件名称是否正确 endl; return -1; } Mat gray; cvtColor(img, gray, COLOR_BGR2GRAY); Mat img_B, img_B_V, gray_B, gray_B_V, gray_T, gray_T_V, gray_TRUNC; //彩色图像二值化 threshold(img, img_B, 125, 255, THRESH_BINARY); threshold(img, img_B_V, 125, 255, THRESH_BINARY_INV); imshow(img_B, img_B); imshow(img_B_V, img_B_V); //灰度图BINARY二值化 threshold(gray, gray_B, 125, 255, THRESH_BINARY); threshold(gray, gray_B_V, 125, 255, THRESH_BINARY_INV); imshow(gray_B, gray_B); imshow(gray_B_V, gray_B_V); //灰度图像TOZERO变换 threshold(gray, gray_T, 125, 255, THRESH_TOZERO); threshold(gray, gray_T_V, 125, 255, THRESH_TOZERO_INV); imshow(gray_T, gray_T); imshow(gray_T_V, gray_T_V); //灰度图像TRUNC变换 threshold(gray, gray_TRUNC, 125, 255, THRESH_TRUNC); imshow(gray_TRUNC, gray_TRUNC); //灰度图像大津法和三角形法二值化 Mat img_Thr imread(C:/opencv/learnOpenCV/threshold.png, IMREAD_GRAYSCALE); Mat img_Thr_O, img_Thr_T; threshold(img_Thr, img_Thr_O, 100, 255, THRESH_BINARY | THRESH_OTSU); threshold(img_Thr, img_Thr_T, 125, 255, THRESH_BINARY | THRESH_TRIANGLE); imshow(img_Thr, img_Thr); imshow(img_Thr_O, img_Thr_O); imshow(img_Thr_T, img_Thr_T); Mat adaptive_mean, adaptive_gauss; adaptiveThreshold(img_Thr, adaptive_mean, 255, ADAPTIVE_THRESH_MEAN_C, THRESH_BINARY, 55, 0); adaptiveThreshold(img_Thr, adaptive_gauss, 255, ADAPTIVE_THRESH_GAUSSIAN_C, THRESH_BINARY, 55, 0); imshow(adaptive_mean, adaptive_mean); imshow(adaptive_gauss, adaptive_gauss); waitKey(0); return 0; }