1 #include "MultiTracker.h"
8 using namespace suanzi;
12 static const std::string TAG = "MultiTracker";
13 static const cv::Size PREFERRED_SIZE = Size(64, 128);
15 static const double MaxCost = 100000;
16 static const int MaxPatch = 5;
18 MultiTracker::MultiTracker(EngineWPtr e)
21 LOG_DEBUG(TAG, "init - loading model.pkl");
22 predictor = PredictorWrapper::create("./python", "./python/model.pkl");
24 this->descriptor = {Size(64, 128), Size(16, 16), Size(8, 8), Size(8, 8), 9};
27 MultiTracker::~MultiTracker()
33 static std::vector<double> similarity(const PatchPtr p1, const PatchPtr p2)
35 std::vector<double> feature;
36 cv::Mat im1(PREFERRED_SIZE, p1->image_crop.type());
37 cv::Mat im2(PREFERRED_SIZE, p2->image_crop.type());
38 cv::resize(p1->image_crop, im1, im1.size());
39 cv::resize(p2->image_crop, im2, im2.size());
41 cv::matchTemplate(im1, im2, result, CV_TM_CCOEFF_NORMED);
42 feature.push_back(result.at<double>(0, 0));
43 cv::matchTemplate(im1, im2, result, CV_TM_CCORR_NORMED);
44 feature.push_back(result.at<double>(0, 0));
47 vector<double>& f1_hog = p1->features.first; Mat f1_hue = p1->features.second;
48 vector<double>& f2_hog = p1->features.first; Mat f2_hue = p1->features.second;
49 feature.push_back(distance_cosine(Eigen::Map<Eigen::VectorXd>(f1_hog.data(), f1_hog.size()),
50 Eigen::Map<Eigen::VectorXd>(f2_hog.data(), f2_hog.size())));
51 feature.push_back(distance_euclidean(Eigen::Map<Eigen::VectorXd>(f1_hog.data(), f1_hog.size()),
52 Eigen::Map<Eigen::VectorXd>(f2_hog.data(), f2_hog.size())));
53 feature.push_back(compareHist(f1_hue, f2_hue, HISTCMP_CORREL));
54 feature.push_back(compareHist(f1_hue, f2_hue, HISTCMP_HELLINGER));
56 Detection& d1 = p1->detection;
57 Detection& d2 = p2->detection;
59 double center_distance = sqrt(pow((d1.center_x - d2.center_x), 2) + pow((d1.center_y - d2.center_y), 2));
60 feature.push_back(center_distance / (d1.width + d1.height + d2.width + d2.height) * 4);
62 feature.push_back(calc_iou_ratio(getRectInDetection(d1), getRectInDetection(d2)));
67 double MultiTracker::distance(TrackerPtr tracker, const cv::Mat& image, const Detection& d)
69 PatchPtr patch = createPatch(image, d);
70 std::vector<double> features;
72 std::vector<double> ss;
73 for (auto i : tracker->patches){
74 ss = similarity(i, patch);
75 features.insert(features.end(), ss.begin(), ss.end());
77 double prob = predictor->predict(4, features);
81 static float calc_iou_ratio(const Detection& d1, const Detection& d2)
83 return calc_iou_ratio(getRectInDetection(d1), getRectInDetection(d2));
86 void MultiTracker::update(unsigned int total, const Detection* detections, const Mat& image)
88 int row = trackers.size();
90 Eigen::MatrixXi cost_matrix = Eigen::MatrixXi::Zero(row, col);
91 for (int i = 0; i < row; i++){
92 for (int j = 0; j < col; j++){
93 if (calc_iou_ratio(trackers[i]->detection, detections[j]) < -0.1)
94 cost_matrix(i, j) = MaxCost;
96 cost_matrix(i, j) = distance(trackers[i], image, detections[j]);
100 Eigen::VectorXi tracker_inds, bb_inds;
101 linear_sum_assignment(cost_matrix, tracker_inds, bb_inds);
103 set<TrackerPtr> unmatched_trackers;
104 set<int> unmatch_bbs_indices;
106 for(unsigned int i = 0; i < trackers.size(); i++){
107 if (!(tracker_inds.array() == i).any()){
108 unmatched_trackers.insert(trackers[i]);
111 for (unsigned int j = 0; j < total; j++){
112 if (!(bb_inds.array() == j).any()){
113 unmatch_bbs_indices.insert(j);
117 // handle matched trackers
118 for (unsigned int i = 0; i < tracker_inds.size(); i++){
119 for (int j = 0; j < bb_inds.size(); j++){
120 int rr = tracker_inds(i);
122 TrackerPtr tracker = trackers[rr];
123 const Detection& detect = detections[cc];
124 if (cost_matrix(rr, cc) < MaxCost){
125 tracker->correct(image, detect);
126 tracker->addPatch(createPatch(image, detect));
128 unmatched_trackers.insert(tracker); // failed trackers
129 unmatch_bbs_indices.insert(cc); // filed detection
134 // handle unmatched trackers
135 for (auto t : unmatched_trackers){
136 t->updateState(image);
139 // handle unmatched detections - Create new trackers
140 vector<Person> inPersons;
141 for (auto i : unmatch_bbs_indices){
142 TrackerPtr new_tracker (new Tracker(image, detections[i]));
143 new_tracker->addPatch(createPatch(image, detections[i]));
144 this->trackers.push_back(new_tracker);
146 inPersons.push_back(test);
149 // callback and notify engine - persons in
150 if (inPersons.size() > 0){
151 if (auto e = engine.lock()){
152 e->onPersonsIn(inPersons);
156 // Delete lost trackers
157 vector<Person> outPersons;
158 for (auto it = trackers.begin(); it < trackers.end(); it++){
159 if ((*it)->status == TrackerStatus::Delete){
161 outPersons.push_back(test);
166 // callback and notify engine - persons out
167 if (outPersons.size() > 0){
168 if (auto e = engine.lock()){
169 e->onPersonsOut(outPersons);
174 static cv::Mat image_crop(const cv::Mat& image, const Detection& bb)
176 return image(getRectInDetection(bb));
179 PatchPtr MultiTracker::createPatch(const Mat& image, const Detection& detect)
181 PatchPtr patch(new Patch());
183 // calculate hog descriptors, size is 3780
185 im = image_crop(image, detect);
186 resize(im, im2, PREFERRED_SIZE);
187 vector<float> feature_hog;
188 this->descriptor.compute(im2, feature_hog);
190 // calculate histogram, size is (64 x 45)
192 cvtColor(im, hsv, COLOR_BGR2HSV);
193 int channels[] = {0, 1};
194 int histSize[] = {45, 64};
195 float hranges[] = {0, 180};
196 float sranges[] = {0, 256};
197 const float* ranges[] = {hranges, sranges};
198 calcHist(&hsv, 1, channels, Mat(), hist, 2, histSize, ranges, true, false);
200 patch->image_crop = im.clone();
201 patch->detection = detect;
202 std::vector<double> feature_hog_double (feature_hog.begin(), feature_hog.end()); // convert to double
203 patch->features = std::make_pair(feature_hog_double, hist);