Task02 Анна Патрушева ИТМО

README.md

@@ -57,4 +57,4 @@
 
 - Если все хорошо, за выполненное задание дается **8 баллов**
 - **3 доп. балла** можно получить, если при оценке матрицы гомографии реализовать метод **A contrario RANSAC**, не требующий на вход порога (см. homography.cpp:166)
-- **1 доп. балл** можно получить, если реализовать Brute-force матчер на GPU. Для включения его в тестах см. **ENABLE_GPU_BRUTEFORCE_MATCHER** в ```test/test_matching.cpp```
+- **1 доп. балл** можно получить, если реализовать Brute-force матчер на GPU. Для включения его в тестах см. **ENABLE_GPU_BRUTEFORCE_MATCHER** в ```test/test_matching.cpp```

src/phg/matching/descriptor_matcher.cpp

@@ -7,8 +7,17 @@ void phg::DescriptorMatcher::filterMatchesRatioTest(const std::vector<std::vecto
                                                     std::vector<cv::DMatch> &filtered_matches)
 {
     filtered_matches.clear();
+    const float threshold = 0.7f;
 
-    throw std::runtime_error("not implemented yet");
+    for (auto& match: matches) {
+        if (match.size() < 2) {
+            continue;
+        }
+        if (match[0].distance < match[1].distance * threshold * threshold) {
+            filtered_matches.push_back(match[0]);
+        }
+    }
+//    throw std::runtime_error("not implemented yet");
 }
 
 
@@ -35,42 +44,71 @@ void phg::DescriptorMatcher::filterMatchesClusters(const std::vector<cv::DMatch>
         points_query.at<cv::Point2f>(i) = keypoints_query[matches[i].queryIdx].pt;
         points_train.at<cv::Point2f>(i) = keypoints_train[matches[i].trainIdx].pt;
     }
-//
-//    // размерность всего 2, так что точное KD-дерево
-//    std::shared_ptr<cv::flann::IndexParams> index_params = flannKdTreeIndexParams(TODO);
-//    std::shared_ptr<cv::flann::SearchParams> search_params = flannKsTreeSearchParams(TODO);
-//
-//    std::shared_ptr<cv::flann::Index> index_query = flannKdTreeIndex(points_query, index_params);
-//    std::shared_ptr<cv::flann::Index> index_train = flannKdTreeIndex(points_train, index_params);
-//
-//    // для каждой точки найти total neighbors ближайших соседей
-//    cv::Mat indices_query(n_matches, total_neighbours, CV_32SC1);
-//    cv::Mat distances2_query(n_matches, total_neighbours, CV_32FC1);
-//    cv::Mat indices_train(n_matches, total_neighbours, CV_32SC1);
-//    cv::Mat distances2_train(n_matches, total_neighbours, CV_32FC1);
-//
-//    index_query->knnSearch(points_query, indices_query, distances2_query, total_neighbours, *search_params);
-//    index_train->knnSearch(points_train, indices_train, distances2_train, total_neighbours, *search_params);
-//
-//    // оценить радиус поиска для каждой картинки
-//    // NB: radius2_query, radius2_train: квадраты радиуса!
-//    float radius2_query, radius2_train;
-//    {
-//        std::vector<double> max_dists2_query(n_matches);
-//        std::vector<double> max_dists2_train(n_matches);
-//        for (int i = 0; i < n_matches; ++i) {
-//            max_dists2_query[i] = distances2_query.at<float>(i, total_neighbours - 1);
-//            max_dists2_train[i] = distances2_train.at<float>(i, total_neighbours - 1);
-//        }
-//
-//        int median_pos = n_matches / 2;
-//        std::nth_element(max_dists2_query.begin(), max_dists2_query.begin() + median_pos, max_dists2_query.end());
-//        std::nth_element(max_dists2_train.begin(), max_dists2_train.begin() + median_pos, max_dists2_train.end());
-//
-//        radius2_query = max_dists2_query[median_pos] * radius_limit_scale * radius_limit_scale;
-//        radius2_train = max_dists2_train[median_pos] * radius_limit_scale * radius_limit_scale;
-//    }
-//
-//    метч остается, если левое и правое множества первых total_neighbors соседей в радиусах поиска(radius2_query, radius2_train) имеют как минимум consistent_matches общих элементов
-//    // TODO заполнить filtered_matches
+
+    // размерность всего 2, так что точное KD-дерево
+    // std::shared_ptr<cv::flann::IndexParams> index_params = flannKdTreeIndexParams(DONE);
+    // std::shared_ptr<cv::flann::SearchParams> search_params = flannKsTreeSearchParams(DONE);
+
+    std::shared_ptr<cv::flann::IndexParams> index_params = flannKdTreeIndexParams(1);
+    std::shared_ptr<cv::flann::SearchParams> search_params = flannKsTreeSearchParams(10);
+
+    std::shared_ptr<cv::flann::Index> index_query = flannKdTreeIndex(points_query, index_params);
+    std::shared_ptr<cv::flann::Index> index_train = flannKdTreeIndex(points_train, index_params);
+
+    // для каждой точки найти total neighbors ближайших соседей
+    cv::Mat indices_query(n_matches, total_neighbours, CV_32SC1);
+    cv::Mat distances2_query(n_matches, total_neighbours, CV_32FC1);
+    cv::Mat indices_train(n_matches, total_neighbours, CV_32SC1);
+    cv::Mat distances2_train(n_matches, total_neighbours, CV_32FC1);
+
+    index_query->knnSearch(points_query, indices_query, distances2_query, total_neighbours, *search_params);
+    index_train->knnSearch(points_train, indices_train, distances2_train, total_neighbours, *search_params);
+
+    // оценить радиус поиска для каждой картинки
+    // NB: radius2_query, radius2_train: квадраты радиуса!
+    float radius2_query, radius2_train;
+    {
+        std::vector<double> max_dists2_query(n_matches);
+        std::vector<double> max_dists2_train(n_matches);
+        for (int i = 0; i < n_matches; ++i) {
+            max_dists2_query[i] = distances2_query.at<float>(i, total_neighbours - 1);
+            max_dists2_train[i] = distances2_train.at<float>(i, total_neighbours - 1);
+        }
+
+        int median_pos = n_matches / 2;
+        std::nth_element(max_dists2_query.begin(), max_dists2_query.begin() + median_pos, max_dists2_query.end());
+        std::nth_element(max_dists2_train.begin(), max_dists2_train.begin() + median_pos, max_dists2_train.end());
+
+        radius2_query = max_dists2_query[median_pos] * radius_limit_scale * radius_limit_scale;
+        radius2_train = max_dists2_train[median_pos] * radius_limit_scale * radius_limit_scale;
+    }
+    // метч остается, если левое и правое множества первых total_neighbors соседей в радиусах поиска(radius2_query, radius2_train) имеют как минимум consistent_matches общих элементов
+    // DONE заполнить filtered_matches
+
+    for (size_t i = 0; i < n_matches; ++i) {
+        std::set<int> neighbours_query, neighbours_train;
+        for (size_t j = 0; j < total_neighbours; ++j) {
+            auto dist2_query = distances2_query.at<float>(i, j);
+            auto dist2_train = distances2_train.at<float>(i, j);
+            if (dist2_query <= radius2_query) {
+                neighbours_query.insert(indices_query.at<int>(i, j));
+            }
+            if (dist2_train <= radius2_train) {
+                neighbours_train.insert(indices_train.at<int>(i, j));
+            }
+        }
+
+        int current_matches = 0;
+        for (auto ind: neighbours_query) {
+            if (neighbours_train.find(ind) != neighbours_train.end()) {
+                current_matches++;
+            }
+        }
+
+        if (current_matches >= consistent_matches) {
+            filtered_matches.push_back(matches[i]);
+        }
+
+    }
+
 }

src/phg/matching/flann_matcher.cpp

@@ -6,8 +6,10 @@
 phg::FlannMatcher::FlannMatcher()
 {
     // параметры для приближенного поиска
-//    index_params = flannKdTreeIndexParams(TODO);
-//    search_params = flannKsTreeSearchParams(TODO);
+    //    index_params = flannKdTreeIndexParams(DONE);
+    //    search_params = flannKsTreeSearchParams(DONE);
+    index_params = flannKdTreeIndexParams(4);
+    search_params = flannKsTreeSearchParams(32);
 }
 
 void phg::FlannMatcher::train(const cv::Mat &train_desc)
@@ -17,5 +19,17 @@ void phg::FlannMatcher::train(const cv::Mat &train_desc)
 
 void phg::FlannMatcher::knnMatch(const cv::Mat &query_desc, std::vector<std::vector<cv::DMatch>> &matches, int k) const
 {
-    throw std::runtime_error("not implemented yet");
+
+    cv::Mat index, dist2;
+    flann_index->knnSearch(query_desc, index, dist2, k, *search_params);
+    matches.resize(query_desc.rows);
+    for (int i = 0; i < query_desc.rows; i++) {
+        matches[i].clear();
+        for (int j = 0; j < k; j++) {
+            int id_train = index.at<int>(i, j);
+            float dist = dist2.at<float>(i, j);
+            matches[i].emplace_back(i, id_train, dist);
+        }
+    }
+//    throw std::runtime_error("not implemented yet");
 }

src/phg/sfm/homography.cpp

@@ -84,8 +84,11 @@ namespace {
             double w1 = ws1[i];
 
             // 8 elements of matrix + free term as needed by gauss routine
-//            A.push_back({TODO});
-//            A.push_back({TODO});
+//            A.push_back({DONE});
+//            A.push_back({DONE});
+            A.push_back({0, 0, 0, -x0 * w1, -y0 * w1, -w0 * w1, x0 * y1, y0 * y1, -y1 * w0});
+            A.push_back({x0 * w1, y0 * w1, w0 * w1, 0, 0, 0, -x0 * x1, -y0 * x1, x1 * w0});
+
         }
 
         int res = gauss(A, H);
@@ -168,57 +171,66 @@ namespace {
         // * (простое описание для понимания)
         // * [3] http://ikrisoft.blogspot.com/2015/01/ransac-with-contrario-approach.html
 
-//        const int n_matches = points_lhs.size();
-//
-//        // https://en.wikipedia.org/wiki/Random_sample_consensus#Parameters
-//        const int n_trials = TODO;
-//
-//        const int n_samples = TODO;
-//        uint64_t seed = 1;
-//        const double reprojection_error_threshold_px = 2;
-//
-//        int best_support = 0;
-//        cv::Mat best_H;
-//
-//        std::vector<int> sample;
-//        for (int i_trial = 0; i_trial < n_trials; ++i_trial) {
-//            randomSample(sample, n_matches, n_samples, &seed);
-//
-//            cv::Mat H = estimateHomography4Points(points_lhs[sample[0]], points_lhs[sample[1]], points_lhs[sample[2]], points_lhs[sample[3]],
-//                                                  points_rhs[sample[0]], points_rhs[sample[1]], points_rhs[sample[2]], points_rhs[sample[3]]);
-//
-//            int support = 0;
-//            for (int i_point = 0; i_point < n_matches; ++i_point) {
-//                try {
-//                    cv::Point2d proj = phg::transformPoint(points_lhs[i_point], H);
-//                    if (cv::norm(proj - cv::Point2d(points_rhs[i_point])) < reprojection_error_threshold_px) {
-//                        ++support;
-//                    }
-//                } catch (const std::exception &e)
-//                {
-//                    std::cerr << e.what() << std::endl;
-//                }
-//            }
-//
-//            if (support > best_support) {
-//                best_support = support;
-//                best_H = H;
-//
-//                std::cout << "estimateHomographyRANSAC : support: " << best_support << "/" << n_matches << std::endl;
-//
-//                if (best_support == n_matches) {
-//                    break;
-//                }
-//            }
-//        }
-//
-//        std::cout << "estimateHomographyRANSAC : best support: " << best_support << "/" << n_matches << std::endl;
-//
-//        if (best_support == 0) {
-//            throw std::runtime_error("estimateHomographyRANSAC : failed to estimate homography");
-//        }
-//
-//        return best_H;
+        const int n_matches = points_lhs.size();
+        const int n_samples = 4;
+
+        // https://en.wikipedia.org/wiki/Random_sample_consensus#Parameters
+
+        const double one_inlier_probability = 0.5f;
+        double n_inliers_probability = 1;
+        for (size_t j = 0; j < n_samples; j++) {
+            n_inliers_probability *= one_inlier_probability;
+        }
+        const double fail_probability = 0.03f;
+        const int n_trials = std::log(fail_probability) / std::log(1 - n_inliers_probability);
+
+//        std::cout << "n_trials = " << n_trials << "\n";
+
+        uint64_t seed = 1;
+        const double reprojection_error_threshold_px = 2;
+
+        int best_support = 0;
+        cv::Mat best_H;
+
+        std::vector<int> sample;
+        for (int i_trial = 0; i_trial < n_trials; ++i_trial) {
+            randomSample(sample, n_matches, n_samples, &seed);
+
+            cv::Mat H = estimateHomography4Points(points_lhs[sample[0]], points_lhs[sample[1]], points_lhs[sample[2]], points_lhs[sample[3]],
+                                                  points_rhs[sample[0]], points_rhs[sample[1]], points_rhs[sample[2]], points_rhs[sample[3]]);
+
+            int support = 0;
+            for (int i_point = 0; i_point < n_matches; ++i_point) {
+                try {
+                    cv::Point2d proj = phg::transformPoint(points_lhs[i_point], H);
+                    if (cv::norm(proj - cv::Point2d(points_rhs[i_point])) < reprojection_error_threshold_px) {
+                        ++support;
+                    }
+                } catch (const std::exception &e)
+                {
+                    std::cerr << e.what() << std::endl;
+                }
+            }
+
+            if (support > best_support) {
+                best_support = support;
+                best_H = H;
+
+                std::cout << "estimateHomographyRANSAC : support: " << best_support << "/" << n_matches << std::endl;
+
+                if (best_support == n_matches) {
+                    break;
+                }
+            }
+        }
+
+        std::cout << "estimateHomographyRANSAC : best support: " << best_support << "/" << n_matches << std::endl;
+
+        if (best_support == 0) {
+            throw std::runtime_error("estimateHomographyRANSAC : failed to estimate homography");
+        }
+
+        return best_H;
     }
 
 }
@@ -238,7 +250,16 @@ cv::Mat phg::findHomographyCV(const std::vector<cv::Point2f> &points_lhs, const
 // таким преобразованием внутри занимается функции cv::perspectiveTransform и cv::warpPerspective
 cv::Point2d phg::transformPoint(const cv::Point2d &pt, const cv::Mat &T)
 {
-    throw std::runtime_error("not implemented yet");
+    double x = T.at<double>(0, 0) * pt.x + T.at<double>(0, 1) * pt.y + T.at<double>(0, 2);
+    double y = T.at<double>(1, 0) * pt.x + T.at<double>(1, 1) * pt.y + T.at<double>(1, 2);
+    double w = T.at<double>(2, 0) * pt.x + T.at<double>(2, 1) * pt.y + T.at<double>(2, 2);
+
+    // DONE
+    if (w == 0.0) {
+        throw std::runtime_error("not implemented yet");
+    }
+
+    return {x / w, y / w};
 }
 
 cv::Point2d phg::transformPointCV(const cv::Point2d &pt, const cv::Mat &T) {

src/phg/sfm/panorama_stitcher.cpp

@@ -3,6 +3,7 @@
 
 #include <libutils/bbox2.h>
 #include <iostream>
+#include <queue>
 
 /*
  * imgs - список картинок
@@ -23,7 +24,43 @@ cv::Mat phg::stitchPanorama(const std::vector<cv::Mat> &imgs,
     {
         // здесь надо посчитать вектор Hs
         // при этом можно обойтись n_images - 1 вызовами функтора homography_builder
-        throw std::runtime_error("not implemented yet");
+
+        int root = -1;
+        for (int v = 0; v < n_images; v++) {
+            if (parent[v] == -1) {
+                root = v;
+            }
+        }
+        std::vector<std::vector<int>> ch(n_images, std::vector<int>());
+        for (int v = 0; v < n_images; v++) {
+            if (parent[v] != -1) {
+                ch[parent[v]].push_back(v);
+            }
+        }
+
+        std::vector<bool> used(n_images, false);
+        used[root] = 1;
+        Hs[root] = cv::Mat::eye(3, 3, CV_64F);
+
+        std::queue<int> q;
+        q.push(root);
+
+        while (!q.empty()) {
+            int v = q.front();
+            q.pop();
+            for (auto& u: ch[v]) {
+                if (!used[u]) {
+                    used[u] = 1;
+                    q.push(u);
+                    cv::Mat H = homography_builder(imgs[u], imgs[v]);
+                    Hs[u] = Hs[v] * H;
+                }
+            }
+        }
+
+
+        // DONE
+        // throw std::runtime_error("not implemented yet");
     }
 
     bbox2<double, cv::Point2d> bbox;