opencv 图像拼接

详细的图像拼接实例注释，但是觉得这个代码整体比较乱，接下来自己会整理一个更加有序的代码。
代码和数据可见
完整的代码和数据请见：代码数据链接

#include 
#include 
#include 
#include "opencv2/opencv_modules.hpp"
#include 
#include "opencv2/imgcodecs.hpp"
#include "opencv2/highgui.hpp"
#include "opencv2/stitching/detail/autocalib.hpp"
#include "opencv2/stitching/detail/blenders.hpp"
#include "opencv2/stitching/detail/timelapsers.hpp"
#include "opencv2/stitching/detail/camera.hpp"
#include "opencv2/stitching/detail/exposure_compensate.hpp"
#include "opencv2/stitching/detail/matchers.hpp"
#include "opencv2/stitching/detail/motion_estimators.hpp"
#include "opencv2/stitching/detail/seam_finders.hpp"
#include "opencv2/stitching/detail/warpers.hpp"
#include "opencv2/stitching/warpers.hpp"
#include 


#define ENABLE_LOG 1
#define LOG(msg) std::cout << msg
#define LOGLN(msg) std::cout << msg << std::endl

using namespace std;
using namespace cv;
using namespace cv::detail;

static void printUsage()
{
    cout <<
        "Rotation model images stitcher.\n\n"
        "stitching_detailed img1 img2 [...imgN] [flags]\n\n"
        "Flags:\n"
        "  --preview\n"
        "      Run stitching in the preview mode. Works faster than usual mode,\n"
        "      but output image will have lower resolution.\n"
        "  --try_cuda (yes|no)\n"
        "      Try to use CUDA. The default value is 'no'. All default values\n"
        "      are for CPU mode.\n"
        "\nMotion Estimation Flags:\n"
        "  --work_megapix \n"
        "      Resolution for image registration step. The default is 0.6 Mpx.\n"
        "  --features (surf|orb)\n"
        "      Type of features used for images matching. The default is surf.\n"
        "  --matcher (homography|affine)\n"
        "      Matcher used for pairwise image matching.\n"
        "  --estimator (homography|affine)\n"
        "      Type of estimator used for transformation estimation.\n"
        "  --match_conf \n"
        "      Confidence for feature matching step. The default is 0.65 for surf and 0.3 for orb.\n"
        "  --conf_thresh \n"
        "      Threshold for two images are from the same panorama confidence.\n"
        "      The default is 1.0.\n"
        "  --ba (no|reproj|ray|affine)\n"
        "      Bundle adjustment cost function. The default is ray.\n"
        "  --ba_refine_mask (mask)\n"
        "      Set refinement mask for bundle adjustment. It looks like 'x_xxx',\n"
        "      where 'x' means refine respective parameter and '_' means don't\n"
        "      refine one, and has the following format:\n"
        "      . The default mask is 'xxxxx'. If bundle\n"
        "      adjustment doesn't support estimation of selected parameter then\n"
        "      the respective flag is ignored.\n"
        "  --wave_correct (no|horiz|vert)\n"
        "      Perform wave effect correction. The default is 'horiz'.\n"
        "  --save_graph \n"
        "      Save matches graph represented in DOT language to  file.\n"
        "      Labels description: Nm is number of matches, Ni is number of inliers,\n"
        "      C is confidence.\n"
        "\nCompositing Flags:\n"
        "  --warp (affine|plane|cylindrical|spherical|fisheye|stereographic|compressedPlaneA2B1|compressedPlaneA1.5B1|compressedPlanePortraitA2B1|compressedPlanePortraitA1.5B1|paniniA2B1|paniniA1.5B1|paniniPortraitA2B1|paniniPortraitA1.5B1|mercator|transverseMercator)\n"
        "      Warp surface type. The default is 'spherical'.\n"
        "  --seam_megapix \n"
        "      Resolution for seam estimation step. The default is 0.1 Mpx.\n"
        "  --seam (no|voronoi|gc_color|gc_colorgrad)\n"
        "      Seam estimation method. The default is 'gc_color'.\n"
        "  --compose_megapix \n"
        "      Resolution for compositing step. Use -1 for original resolution.\n"
        "      The default is -1.\n"
        "  --expos_comp (no|gain|gain_blocks)\n"
        "      Exposure compensation method. The default is 'gain_blocks'.\n"
        "  --blend (no|feather|multiband)\n"
        "      Blending method. The default is 'multiband'.\n"
        "  --blend_strength \n"
        "      Blending strength from [0,100] range. The default is 5.\n"
        "  --output \n"
        "      The default is 'result.jpg'.\n"
        "  --timelapse (as_is|crop) \n"
        "      Output warped images separately as frames of a time lapse movie, with 'fixed_' prepended to input file names.\n"
        "  --rangewidth \n"
        "      uses range_width to limit number of images to match with.\n";
}


// Default command line args
vector img_names;
bool preview = false;  // 使用preview将会加快运算速度，但同时降低输出图像分辨率
bool try_cuda = false; // 是否使用CUDA加速
double work_megapix = 0.6; //图像匹配步骤的分辨率？？？？
double seam_megapix = 0.1; // 拼缝图像分辨率？？？
double compose_megapix = -1; //曝光补偿时候分辨率 -1表示使用原始分辨率
float conf_thresh = 1.f; //两幅图来自同一个全景图的置信度
string features_type = "surf"; //特征点选取 SURF或者ORB
string matcher_type = "homography"; //匹配方法 affine或者homography(射影变换)
string estimator_type = "homography"; //预测参数值方法 affine或者homography(射影变换)
string ba_cost_func = "ray"; //光束平差法损失函数 (no|reproj|ray|affine)
string ba_refine_mask = "xxxxx"; //当使用ray作为光束平差法损失函数时，需要初始化setRefinementMask（表示需要精确化的相机内参数矩阵K的掩码矩阵)

bool do_wave_correct = true; //波形矫正 (no|horiz|vert)，默认为水平方向
WaveCorrectKind wave_correct = detail::WAVE_CORRECT_HORIZ;

bool save_graph = true; //存储匹配对文件名 label中Nm为匹配数量，Ni为内点数，C为置信度
std::string save_graph_to;
string warp_type = "spherical";// warp (affine|plane|cylindrical|spherical|fisheye|stereographic|等等
int expos_comp_type = ExposureCompensator::GAIN_BLOCKS;//增益补偿(no|gain|gain_blocks)
float match_conf = 0.3f; //匹配点对的置信度
string seam_find_type = "gc_color";
int blend_type = Blender::MULTI_BAND;//图像融合方法 blend (no|feather|multiband)
int timelapse_type = Timelapser::AS_IS;
float blend_strength = 5; // 这里好像是跟融合 下采样之类的数量有关
string result_name = "result.jpg";
bool timelapse = false; // timelapse (as_is|crop)
int range_width = -1; // 限制图像匹配的数量


static int parseCmdArgs(int argc, char** argv)
{
    if (argc == 1)
    {
        printUsage();
        return -1;
    }
    for (int i = 1; i < argc; ++i)
    {
        if (string(argv[i]) == "--help" || string(argv[i]) == "/?")
        {
            printUsage();
            return -1;
        }
        else if (string(argv[i]) == "--preview")
        {
            preview = true;
        }
        else if (string(argv[i]) == "--try_cuda")
        {
            if (string(argv[i + 1]) == "no")
                try_cuda = false;
            else if (string(argv[i + 1]) == "yes")
                try_cuda = true;
            else
            {
                cout << "Bad --try_cuda flag value\n";
                return -1;
            }
            i++;
        }
        else if (string(argv[i]) == "--work_megapix")
        {
            work_megapix = atof(argv[i + 1]);
            i++;
        }
        else if (string(argv[i]) == "--seam_megapix")
        {
            seam_megapix = atof(argv[i + 1]);
            i++;
        }
        else if (string(argv[i]) == "--compose_megapix")
        {
            compose_megapix = atof(argv[i + 1]);
            i++;
        }
        else if (string(argv[i]) == "--result")
        {
            result_name = argv[i + 1];
            i++;
        }
        else if (string(argv[i]) == "--features")
        {
            features_type = argv[i + 1];
            if (features_type == "orb")
                match_conf = 0.3f;
            i++;
        }
        else if (string(argv[i]) == "--matcher")
        {
            if (string(argv[i + 1]) == "homography" || string(argv[i + 1]) == "affine")
                matcher_type = argv[i + 1];
            else
            {
                cout << "Bad --matcher flag value\n";
                return -1;
            }
            i++;
        }
        else if (string(argv[i]) == "--estimator")
        {
            if (string(argv[i + 1]) == "homography" || string(argv[i + 1]) == "affine")
                estimator_type = argv[i + 1];
            else
            {
                cout << "Bad --estimator flag value\n";
                return -1;
            }
            i++;
        }
        else if (string(argv[i]) == "--match_conf")
        {
            match_conf = static_cast(atof(argv[i + 1]));
            i++;
        }
        else if (string(argv[i]) == "--conf_thresh")
        {
            conf_thresh = static_cast(atof(argv[i + 1]));
            i++;
        }
        else if (string(argv[i]) == "--ba")
        {
            ba_cost_func = argv[i + 1];
            i++;
        }
        else if (string(argv[i]) == "--ba_refine_mask")
        {
            ba_refine_mask = argv[i + 1];
            if (ba_refine_mask.size() != 5)
            {
                cout << "Incorrect refinement mask length.\n";
                return -1;
            }
            i++;
        }
        else if (string(argv[i]) == "--wave_correct")
        {
            if (string(argv[i + 1]) == "no")
                do_wave_correct = false;
            else if (string(argv[i + 1]) == "horiz")
            {
                do_wave_correct = true;
                wave_correct = detail::WAVE_CORRECT_HORIZ;
            }
            else if (string(argv[i + 1]) == "vert")
            {
                do_wave_correct = true;
                wave_correct = detail::WAVE_CORRECT_VERT;
            }
            else
            {
                cout << "Bad --wave_correct flag value\n";
                return -1;
            }
            i++;
        }
        else if (string(argv[i]) == "--save_graph")
        {
            save_graph = true;
            save_graph_to = argv[i + 1];
            i++;
        }
        else if (string(argv[i]) == "--warp")
        {
            warp_type = string(argv[i + 1]);
            i++;
        }
        else if (string(argv[i]) == "--expos_comp")
        {
            if (string(argv[i + 1]) == "no")
                expos_comp_type = ExposureCompensator::NO;
            else if (string(argv[i + 1]) == "gain")
                expos_comp_type = ExposureCompensator::GAIN;
            else if (string(argv[i + 1]) == "gain_blocks")
                expos_comp_type = ExposureCompensator::GAIN_BLOCKS;
            else
            {
                cout << "Bad exposure compensation method\n";
                return -1;
            }
            i++;
        }
        else if (string(argv[i]) == "--seam")
        {
            if (string(argv[i + 1]) == "no" ||
                string(argv[i + 1]) == "voronoi" ||
                string(argv[i + 1]) == "gc_color" ||
                string(argv[i + 1]) == "gc_colorgrad" ||
                string(argv[i + 1]) == "dp_color" ||
                string(argv[i + 1]) == "dp_colorgrad")
                seam_find_type = argv[i + 1];
            else
            {
                cout << "Bad seam finding method\n";
                return -1;
            }
            i++;
        }
        else if (string(argv[i]) == "--blend")
        {
            if (string(argv[i + 1]) == "no")
                blend_type = Blender::NO;
            else if (string(argv[i + 1]) == "feather")
                blend_type = Blender::FEATHER;
            else if (string(argv[i + 1]) == "multiband")
                blend_type = Blender::MULTI_BAND;
            else
            {
                cout << "Bad blending method\n";
                return -1;
            }
            i++;
        }
        else if (string(argv[i]) == "--timelapse")
        {
            timelapse = true;

            if (string(argv[i + 1]) == "as_is")
                timelapse_type = Timelapser::AS_IS;
            else if (string(argv[i + 1]) == "crop")
                timelapse_type = Timelapser::CROP;
            else
            {
                cout << "Bad timelapse method\n";
                return -1;
            }
            i++;
        }
        else if (string(argv[i]) == "--rangewidth")
        {
            range_width = atoi(argv[i + 1]);
            i++;
        }
        else if (string(argv[i]) == "--blend_strength")
        {
            blend_strength = static_cast(atof(argv[i + 1]));
            i++;
        }
        else if (string(argv[i]) == "--output")
        {
            result_name = argv[i + 1];
            i++;
        }
        else
            img_names.push_back(argv[i]);
    }
    if (preview)
    {
        compose_megapix = 0.6;
    }
    return 0;
}


int main(int argc, char* argv[])
{
#if ENABLE_LOG
    int64 app_start_time = getTickCount(); // 统计时间
#endif

#if 0
    cv::setBreakOnError(true);
#endif

    int retval = parseCmdArgs(argc, argv);
    if (retval)
        return retval;

    // Check if have enough images
    int num_images = static_cast(img_names.size());
    if (num_images < 2)
    {
        LOGLN("Need more images"); // LOGIN 和 LOG都使用define语句定义过了 就是一个cout
        return -1;
    }

    double work_scale = 1, seam_scale = 1, compose_scale = 1;
    bool is_work_scale_set = false, is_seam_scale_set = false, is_compose_scale_set = false;

    LOGLN("Finding features...");
#if ENABLE_LOG
    int64 t = getTickCount();
#endif
    // 第一步 寻找特征点 surf 或者 orb特征

    cv::initModule_nonfree();

    Ptr finder; //Ptr是opencv中智能指针
    if (features_type == "surf")
    {
#ifdef HAVE_OPENCV_XFEATURES2D
        if (try_cuda && cuda::getCudaEnabledDeviceCount() > 0)
            finder = makePtr();
        else
#endif
            finder = makePtr();
    }
    else if (features_type == "orb")
    {
        finder = makePtr();
    }
    else
    {
        cout << "Unknown 2D features type: '" << features_type << "'.\n";
        return -1;
    }

    Mat full_img, img;
    vector features(num_images);
    vector images(num_images);
    vector full_img_sizes(num_images); //存储每一张图像的大小
    double seam_work_aspect = 1;

    for (int i = 0; i < num_images; ++i)
    {
        full_img = imread(img_names[i]);
        full_img_sizes[i] = full_img.size();

        if (full_img.empty())
        {
            LOGLN("Can't open image " << img_names[i]);
            return -1;
        }
        if (work_megapix < 0)
        {
            img = full_img;
            work_scale = 1;
            is_work_scale_set = true;
        }
        else
        {
            if (!is_work_scale_set)
            {
                work_scale = min(1.0, sqrt(work_megapix * 1e6 / full_img.size().area()));
                is_work_scale_set = true;
            }
            resize(full_img, img, Size(), work_scale, work_scale, INTER_LINEAR_EXACT);//work_scale代表长宽方向缩放的尺度,则配准时图像分辨率为0.6Mpix
        }
        if (!is_seam_scale_set)
        {
            seam_scale = min(1.0, sqrt(seam_megapix * 1e6 / full_img.size().area()));
            seam_work_aspect = seam_scale / work_scale;//类似的定义了拼缝的分辨率
            is_seam_scale_set = true;
        }

        (*finder)(img, features[i]);
        features[i].img_idx = i;//讲匹配结果存储在features
        vector img_feature(num_images);//定义一个图像存储特征

        LOGLN("Features in image #" << i+1 << ": " << features[i].keypoints.size());
//        drawKeypoints(img, featurs[i],img_feature[i], Scalar::all(-1));//绘制特征点
//        namedWindow("feature");
//        imshow("feature",img_feature[i]);
//        waitKey(500);


        resize(full_img, img, Size(), seam_scale, seam_scale, INTER_LINEAR_EXACT);//这里已经修改为了拼缝的分辨率
        images[i] = img.clone();
    }

    //释放内存
    finder->collectGarbage();
    full_img.release();
    img.release();

    LOGLN("Finding features, time: " << ((getTickCount() - t) / getTickFrequency()) << " sec");

    LOG("Pairwise matching");
#if ENABLE_LOG
    t = getTickCount();
#endif
    vector pairwise_matches;
    Ptr matcher;//智能指针
    if (matcher_type == "affine")
        matcher = makePtr(false, try_cuda, match_conf);
        //使用2NN方法进行特征点匹配，并且当描述子的比值大于阈值认为是正确匹配
        //lab/lcd<1-match_conf则认为ab是正确匹配，在此之前如果寻找到匹配点数量小于2，则退出
    else if (range_width==-1)// 每幅图允许匹配的数量 可能是考虑到了投影变换需要计算参数更多
        matcher = makePtr(try_cuda, match_conf);// makePtr 相当于Ptr
    else
        matcher = makePtr(range_width, try_cuda, match_conf);

    (*matcher)(features, pairwise_matches);
    matcher->collectGarbage();
    LOGLN("Pairwise matching, time: " << ((getTickCount() - t) / getTickFrequency()) << " sec");



    // Check if we should save matches graph，是否存储匹配对
    if (save_graph)
    {
        LOGLN("Saving matches graph...");
        ofstream f(save_graph_to.c_str());
        f << matchesGraphAsString(img_names, pairwise_matches, conf_thresh);
    }

    // Leave only images we are sure are from the same panorama
    // 这里是否对features和pairwist_matches进行了更新？？？？
    // 这里给出了源码 可以看出更新了feature和pairwise_matches https://www.cnblogs.com/jsxyhelu/p/6810964.html
    vector indices = leaveBiggestComponent(features, pairwise_matches, conf_thresh);
    //c = ni/(8+3N） 如果这个数大于3，则认为是同一幅图，这一步骤已经集成在函数内部，如果低于阈值，则认为是不能拼接，在这里机构建最大可拼接子集
    vector img_subset;
    vector img_names_subset;
    vector full_img_sizes_subset;// 更新图像集合为全部可拼接图像
    for (size_t i = 0; i < indices.size(); ++i)
    {
        img_names_subset.push_back(img_names[indices[i]]);
        img_subset.push_back(images[indices[i]]);
        full_img_sizes_subset.push_back(full_img_sizes[indices[i]]);
    }

    images = img_subset;
    img_names = img_names_subset;
    full_img_sizes = full_img_sizes_subset;

    // Check if we still have enough images
    num_images = static_cast(img_names.size());
    if (num_images < 2)
    {
        LOGLN("Need more images");
        return -1;
    }

    Ptr estimator;
    if (estimator_type == "affine")
        estimator = makePtr();
    else
        estimator = makePtr();

    vector cameras;//这里的旋转矩阵包括了相机的内参以及旋转和平移向量,这里只是初步预测，后面使用光束平差法进行了细化
    if (!(*estimator)(features, pairwise_matches, cameras))
    {
        cout << "Homography estimation failed.\n";
        return -1;
    }

    for (size_t i = 0; i < cameras.size(); ++i)
    {
        Mat R;
        cameras[i].R.convertTo(R, CV_32F);// 转换旋转矩阵的数据类型
        cameras[i].R = R;
        LOGLN("Initial camera intrinsics #" << indices[i]+1 << ":\nK:\n" << cameras[i].K() << "\nR:\n" << cameras[i].R);
    }

    Ptr adjuster;
    if (ba_cost_func == "reproj") adjuster = makePtr();
    else if (ba_cost_func == "ray") adjuster = makePtr();
    else if (ba_cost_func == "affine") adjuster = makePtr();
    else if (ba_cost_func == "no") adjuster = makePtr();
    else
    {
        cout << "Unknown bundle adjustment cost function: '" << ba_cost_func << "'.\n";
        return -1;
    }
    //这个我觉得没什么用 前面已经通过leaveBiggestComponent求出了最大子集
    adjuster->setConfThresh(conf_thresh);


    //当使用ray作为光束平差法损失函数时，需要初始化setRefinementMask（表示需要精确化的相机内参数矩阵K的掩码矩阵)
    Mat_ refine_mask = Mat::zeros(3, 3, CV_8U);
    if (ba_refine_mask[0] == 'x') refine_mask(0,0) = 1;
    if (ba_refine_mask[1] == 'x') refine_mask(0,1) = 1;
    if (ba_refine_mask[2] == 'x') refine_mask(0,2) = 1;
    if (ba_refine_mask[3] == 'x') refine_mask(1,1) = 1;
    if (ba_refine_mask[4] == 'x') refine_mask(1,2) = 1;
    adjuster->setRefinementMask(refine_mask);
    if (!(*adjuster)(features, pairwise_matches, cameras))
    {
        cout << "Camera parameters adjusting failed.\n";
        return -1;
    }

    // Find median focal length， 这里的focal取得是中值，也可以取平均值

    vector focals;
    for (size_t i = 0; i < cameras.size(); ++i)
    {
        LOGLN("Camera #" << indices[i]+1 << ":\nK:\n" << cameras[i].K() << "\nR:\n" << cameras[i].R);
        focals.push_back(cameras[i].focal);
    }

    sort(focals.begin(), focals.end());


    float warped_image_scale;
    if (focals.size() % 2 == 1)
        warped_image_scale = static_cast(focals[focals.size() / 2]);
    else
        warped_image_scale = static_cast(focals[focals.size() / 2 - 1] + focals[focals.size() / 2]) * 0.5f;
    // 类似论文中的up vector
    if (do_wave_correct)
    {
        vector rmats;
        for (size_t i = 0; i < cameras.size(); ++i)
            rmats.push_back(cameras[i].R.clone());
        waveCorrect(rmats, wave_correct);
        for (size_t i = 0; i < cameras.size(); ++i)
            cameras[i].R = rmats[i];
    }


    // 由于在拍摄时候，图像位于不同的平面，如果直接拼接的话，会破坏是视觉场的一致性，所以要将其映射到平面上
    LOGLN("Warping images (auxiliary)... ");
#if ENABLE_LOG
    t = getTickCount();
#endif

    vector corners(num_images); // 映射之后图像左上角坐标
    vector masks_warped(num_images); // 映射图像后的掩码
    vector images_warped(num_images); // 映射变换后图像
    vector sizes(num_images); // 映射后图像尺寸
    vector masks(num_images); // 原图尺寸

    // Preapre images masks
    for (int i = 0; i < num_images; ++i)
    {
        masks[i].create(images[i].size(), CV_8U);
        masks[i].setTo(Scalar::all(255));//定义原图中所有部分均使用
    }


    // Warp images and their masks
    // 将最终的图像进行映射变换，最终是在平面 椭圆还是其他
    Ptr warper_creator;
#ifdef HAVE_OPENCV_CUDAWARPING
    if (try_cuda && cuda::getCudaEnabledDeviceCount() > 0)
    {
        if (warp_type == "plane")
            warper_creator = makePtr();
        else if (warp_type == "cylindrical")
            warper_creator = makePtr();
        else if (warp_type == "spherical")
            warper_creator = makePtr();
    }
    else
#endif
    {
        if (warp_type == "plane")
            warper_creator = makePtr();
        else if (warp_type == "affine")
            warper_creator = makePtr();
        else if (warp_type == "cylindrical")
            warper_creator = makePtr();
        else if (warp_type == "spherical")
            warper_creator = makePtr();
        else if (warp_type == "fisheye")
            warper_creator = makePtr();
        else if (warp_type == "stereographic")
            warper_creator = makePtr();
        else if (warp_type == "compressedPlaneA2B1")
            warper_creator = makePtr(2.0f, 1.0f);
        else if (warp_type == "compressedPlaneA1.5B1")
            warper_creator = makePtr(1.5f, 1.0f);
        else if (warp_type == "compressedPlanePortraitA2B1")
            warper_creator = makePtr(2.0f, 1.0f);
        else if (warp_type == "compressedPlanePortraitA1.5B1")
            warper_creator = makePtr(1.5f, 1.0f);
        else if (warp_type == "paniniA2B1")
            warper_creator = makePtr(2.0f, 1.0f);
        else if (warp_type == "paniniA1.5B1")
            warper_creator = makePtr(1.5f, 1.0f);
        else if (warp_type == "paniniPortraitA2B1")
            warper_creator = makePtr(2.0f, 1.0f);
        else if (warp_type == "paniniPortraitA1.5B1")
            warper_creator = makePtr(1.5f, 1.0f);
        else if (warp_type == "mercator")
            warper_creator = makePtr();
        else if (warp_type == "transverseMercator")
            warper_creator = makePtr();
    }

    if (!warper_creator)
    {
        cout << "Can't create the following warper '" << warp_type << "'\n";
        return 1;
    }
    // 参数数量视映射情况而定,设置映射的尺寸为焦距，这里是因为定义了拼缝，所有乘以了拼缝；
    Ptr warper = warper_creator->create(static_cast(warped_image_scale * seam_work_aspect));

    for (int i = 0; i < num_images; ++i)
    {
        Mat_ K;// K为相机内参
        cameras[i].K().convertTo(K, CV_32F);
        float swa = (float)seam_work_aspect;
        K(0,0) *= swa; K(0,2) *= swa;
        K(1,1) *= swa; K(1,2) *= swa;
        // 这里用到了相机的内参和外参，得到了变换后图像左上角坐标和变换后图像
        corners[i] = warper->warp(images[i], K, cameras[i].R, INTER_LINEAR, BORDER_REFLECT, images_warped[i]);
        sizes[i] = images_warped[i].size();// 映射后图像尺寸
        // 得到了映射后的图像掩码
        warper->warp(masks[i], K, cameras[i].R, INTER_NEAREST, BORDER_CONSTANT, masks_warped[i]);
    }

    vector images_warped_f(num_images);
    for (int i = 0; i < num_images; ++i)
        images_warped[i].convertTo(images_warped_f[i], CV_32F);

    LOGLN("Warping images, time: " << ((getTickCount() - t) / getTickFrequency()) << " sec");

    Ptr compensator = ExposureCompensator::createDefault(expos_comp_type);
    compensator->feed(corners, images_warped, masks_warped);

    // 定义拼缝
    Ptr seam_finder;
    if (seam_find_type == "no")
        seam_finder = makePtr();
    else if (seam_find_type == "voronoi")
        seam_finder = makePtr();
    else if (seam_find_type == "gc_color")
    {
#ifdef HAVE_OPENCV_CUDALEGACY
        if (try_cuda && cuda::getCudaEnabledDeviceCount() > 0)
            seam_finder = makePtr(GraphCutSeamFinderBase::COST_COLOR);
        else
#endif
            seam_finder = makePtr(GraphCutSeamFinderBase::COST_COLOR);
    }
    else if (seam_find_type == "gc_colorgrad")
    {
#ifdef HAVE_OPENCV_CUDALEGACY
        if (try_cuda && cuda::getCudaEnabledDeviceCount() > 0)
            seam_finder = makePtr(GraphCutSeamFinderBase::COST_COLOR_GRAD);
        else
#endif
            seam_finder = makePtr(GraphCutSeamFinderBase::COST_COLOR_GRAD);
    }
    else if (seam_find_type == "dp_color")
        seam_finder = makePtr(DpSeamFinder::COLOR);
    else if (seam_find_type == "dp_colorgrad")
        seam_finder = makePtr(DpSeamFinder::COLOR_GRAD);
    if (!seam_finder)
    {
        cout << "Can't create the following seam finder '" << seam_find_type << "'\n";
        return 1;
    }

    // 得到接缝线的掩码图像

    seam_finder->find(images_warped_f, corners, masks_warped);

    // Release unused memory
    images.clear();
    images_warped.clear();
    images_warped_f.clear();
    masks.clear();

    LOGLN("Compositing...");
#if ENABLE_LOG
    t = getTickCount();
#endif


// 进行曝光补偿，这里由于曝光时尺寸发生了变换，因此这里需要对映射后的分辨率进行改变
    Mat img_warped, img_warped_s;
    Mat dilated_mask, seam_mask, mask, mask_warped;
    Ptr blender;
    Ptr timelapser;
    //double compose_seam_aspect = 1;
    double compose_work_aspect = 1;

    for (int img_idx = 0; img_idx < num_images; ++img_idx)
    {
        LOGLN("Compositing image #" << indices[img_idx]+1);

        // Read image and resize it if necessary
        full_img = imread(img_names[img_idx]);
        if (!is_compose_scale_set)
        {
            if (compose_megapix > 0)
                compose_scale = min(1.0, sqrt(compose_megapix * 1e6 / full_img.size().area()));
            is_compose_scale_set = true;

            // Compute relative scales
            //compose_seam_aspect = compose_scale / seam_scale;
            compose_work_aspect = compose_scale / work_scale;

            // Update warped image scale，warped_image_scale是焦距尺寸
            warped_image_scale *= static_cast(compose_work_aspect);
            warper = warper_creator->create(warped_image_scale);

            // Update corners and sizes
            for (int i = 0; i < num_images; ++i)
            {
                // Update intrinsics
                cameras[i].focal *= compose_work_aspect;
                cameras[i].ppx *= compose_work_aspect; //Principal point X
                cameras[i].ppy *= compose_work_aspect;

                // Update corner and size
                Size sz = full_img_sizes[i];
                if (std::abs(compose_scale - 1) > 1e-1)
                {
                    sz.width = cvRound(full_img_sizes[i].width * compose_scale);
                    sz.height = cvRound(full_img_sizes[i].height * compose_scale);
                }

                Mat K;
                cameras[i].K().convertTo(K, CV_32F);
                Rect roi = warper->warpRoi(sz, K, cameras[i].R);// Projected image minimum bounding box
                corners[i] = roi.tl();//左上角坐标
                sizes[i] = roi.size();//尺寸
            }
        }
        if (abs(compose_scale - 1) > 1e-1)
            resize(full_img, img, Size(), compose_scale, compose_scale, INTER_LINEAR_EXACT);
        else
            img = full_img;
        full_img.release();
        Size img_size = img.size();

        Mat K;
        cameras[img_idx].K().convertTo(K, CV_32F);

        // 这里为什么要重新进行映射 单纯是因为尺寸原因？？？
        warper->warp(img, K, cameras[img_idx].R, INTER_LINEAR, BORDER_REFLECT, img_warped);

        // Warp the current image mask
        mask.create(img_size, CV_8U);
        mask.setTo(Scalar::all(255));
        warper->warp(mask, K, cameras[img_idx].R, INTER_NEAREST, BORDER_CONSTANT, mask_warped);

        // Compensate exposure
        compensator->apply(img_idx, corners[img_idx], img_warped, mask_warped);

        img_warped.convertTo(img_warped_s, CV_16S);
        img_warped.release();
        img.release();
        mask.release();

        // 这里来说结构元素是什么 Mat()是什么意思？？？？？
        //在融合的时候，最重要的是在接缝线两侧进行处理，而上一步在寻找接缝线后得到的掩码的边界就是接缝线处，
        // 因此我们还需要在接缝线两侧开辟一块区域用于融合处理，这一处理过程对羽化方法尤为关键
        // 应用膨胀算法缩小掩码面积
        dilate(masks_warped[img_idx], dilated_mask, Mat());
        resize(dilated_mask, seam_mask, mask_warped.size(), 0, 0, INTER_LINEAR_EXACT);

        // 映射变换图的掩码和膨胀后的掩码相“与”，从而使扩展的区域仅仅限于接缝线两侧，其他边界处不受影响

        mask_warped = seam_mask & mask_warped;

        if (!blender && !timelapse)
        {
            blender = Blender::createDefault(blend_type, try_cuda);
            Size dst_sz = resultRoi(corners, sizes).size();
            float blend_width = sqrt(static_cast(dst_sz.area())) * blend_strength / 100.f;
            if (blend_width < 1.f)
                blender = Blender::createDefault(Blender::NO, try_cuda);
            else if (blend_type == Blender::MULTI_BAND)
            {
                MultiBandBlender* mb = dynamic_cast(blender.get());
                //设置频段数，即金字塔层数
                mb->setNumBands(static_cast(ceil(log(blend_width)/log(2.)) - 1.));
                LOGLN("Multi-band blender, number of bands: " << mb->numBands());
            }
            else if (blend_type == Blender::FEATHER)
            {
                FeatherBlender* fb = dynamic_cast(blender.get());
                fb->setSharpness(1.f/blend_width);// 设置羽化度
                LOGLN("Feather blender, sharpness: " << fb->sharpness());
            }
            blender->prepare(corners, sizes);
        }
        else if (!timelapser && timelapse)
        {
            timelapser = Timelapser::createDefault(timelapse_type);
            timelapser->initialize(corners, sizes);
        }

        // Blend the current image
        if (timelapse)
        {
            timelapser->process(img_warped_s, Mat::ones(img_warped_s.size(), CV_8UC1), corners[img_idx]);
            String fixedFileName;
            size_t pos_s = String(img_names[img_idx]).find_last_of("/\\");
            if (pos_s == String::npos)
            {
                fixedFileName = "fixed_" + img_names[img_idx];
            }
            else
            {
                fixedFileName = "fixed_" + String(img_names[img_idx]).substr(pos_s + 1, String(img_names[img_idx]).length() - pos_s);
            }
            imwrite(fixedFileName, timelapser->getDst());
        }
        else
        {
            blender->feed(img_warped_s, mask_warped, corners[img_idx]);
        }
    }

    if (!timelapse)
    {
        Mat result, result_mask;
        blender->blend(result, result_mask);

        LOGLN("Compositing, time: " << ((getTickCount() - t) / getTickFrequency()) << " sec");

        imwrite(result_name, result);
    }

    LOGLN("Finished, total time: " << ((getTickCount() - app_start_time) / getTickFrequency()) << " sec");
    return 0;
}

拼接结果：