Mercurial Hosting > traffic-intelligence
view trajectorymanagement/src/LCSMetric.h @ 1275:9f1711a85c56
added code to make sure TTC is replaced of recomputed and not having a value
author | Nicolas Saunier <nicolas.saunier@polymtl.ca> |
---|---|
date | Wed, 19 Jun 2024 16:11:35 -0400 |
parents | e1e7acef8eab |
children |
line wrap: on
line source
#ifndef LCSMETRIC_H_ #define LCSMETRIC_H_ #include "Metric.h" /** * LCSMetric class. * * The Longest Common Subsequence metric * This class mesures : * 1) the similarity (number of points in "common", ie relative to distance similarity threshold) ; * 2) the normalized distance between two trajectories */ template<typename Tr, typename To> class LCSMetric: public Metric<Tr, To> { public: /** * Constructor. */ LCSMetric() : similarityThreshold(0.0), eps(1e-6) { } /** * Set similarity threshold between two points. * * @param[in] similarityThreshold similarity threshold */ bool setSimilarityThreshold(double similarityThreshold) { if (similarityThreshold >= 0.0) { this->similarityThreshold = similarityThreshold; return true; } return false; } /** * Set machine epsilon. * * @param[in] eps machine epsilon */ bool setEps(double eps) { if (eps >= 0.0) { this->eps = eps; return true; } return false; } /** * Compute similarity between two trajectories. * * @param[in] a input trajectory * @param[in] b input trajectory * @param[out] result distance between two trajectories */ void distance(const Trajectory<Tr> *a, const Trajectory<Tr> *b, To &result, unsigned int nbOfPoints = std::numeric_limits<unsigned int>::max()) { result = To(0); unsigned int LCSS = 0; similarity(a,b, LCSS); unsigned int min_size = min(a->size(),b->size()); result = 1 - double(LCSS/min_size); } /** * Compute similarity between two trajectories. * * @param[in] a input trajectory * @param[in] b input trajectory * @param[out] result similarity between two trajectories */ void similarity(const Trajectory<Tr> *a, const Trajectory<Tr> *b, unsigned int &result) { unsigned int LCS[a->size() + 1][b->size() + 1]; { //initialisation for (unsigned int i = 0; i <= a->size(); ++i) { LCS[i][0] = 0; } for (unsigned int j = 0; j <= b->size(); ++j) { LCS[0][j] = 0; } } { //algorithm for (unsigned int i = 1; i <= a->size(); ++i) { for (unsigned int j = 1; j <= b->size(); ++j) { cv::Point3_<typeof(static_cast<Tr>(a->getPoint(i-1)).x)> p(a->getPoint(i - 1) - b->getPoint(j - 1)); double distance = sqrt(pow(p.x, 2) + pow(p.y, 2) + pow(p.z, 2)); // il faudrait généraliser if (distance <= similarityThreshold + eps) { //a[i] == b[j] LCS[i][j] = LCS[i - 1][j - 1] + 1; } else { LCS[i][j] = std::max(LCS[i - 1][j], LCS[i][j - 1]); } } } } result = LCS[a->size()][b->size()]; } private: /** * Similarity threshold between two points. */ double similarityThreshold; /** * Machine epsilon. */ double eps; }; #endif /* LCSMETRIC_H_ */