425 lines
16 KiB
C++
Executable File
425 lines
16 KiB
C++
Executable File
#include <iostream>
|
|
|
|
#include "filter/Structs.h"
|
|
#include "filter/KLB.h"
|
|
#include "Plotti.h"
|
|
#include "filter/Logic.h"
|
|
#include "Settings.h"
|
|
|
|
#include <sys/types.h>
|
|
#include <sys/stat.h>
|
|
|
|
#include <Indoor/sensors/radio/model/WiFiModelFactory.h>
|
|
#include <Indoor/sensors/radio/model/WiFiModelFactoryImpl.h>
|
|
|
|
|
|
//frank
|
|
//const std::string mapDir = "/mnt/data/workspaces/IPIN2016/IPIN2016/competition/maps/";
|
|
//const std::string dataDir = "/mnt/data/workspaces/IPIN2016/IPIN2016/competition/src/data/";
|
|
|
|
//toni
|
|
const std::string mapDir = "/home/toni/Documents/programme/localization/IndoorMap/maps/";
|
|
//const std::string dataDir = "/home/toni/Documents/programme/localization/DynLag/code/data/";
|
|
const std::string dataDir = "/home/toni/Documents/programme/localization/museum/measurements/shl/";
|
|
const std::string errorDir = dataDir + "results/";
|
|
|
|
/** describes one dataset (map, training, parameter-estimation, ...) */
|
|
struct DataSetup {
|
|
std::string map;
|
|
std::vector<std::string> training;
|
|
std::string wifiParams;
|
|
int minWifiOccurences;
|
|
VAPGrouper::Mode vapMode;
|
|
std::string grid;
|
|
};
|
|
|
|
/** all configured datasets */
|
|
struct Data {
|
|
|
|
DataSetup SecondFloorOnly = {
|
|
|
|
mapDir + "SHL_Stock_2_01.xml",
|
|
|
|
{
|
|
dataDir + "Path1_1.csv",
|
|
dataDir + "Path2_1.csv",
|
|
dataDir + "Path3_1.csv",
|
|
},
|
|
|
|
mapDir + "wifi_fp_all.dat",
|
|
40,
|
|
VAPGrouper::Mode::LAST_MAC_DIGIT_TO_ZERO,
|
|
mapDir + "grid_Stock_2_01.dat"
|
|
};
|
|
|
|
} data;
|
|
|
|
|
|
Floorplan::IndoorMap* MyState::map;
|
|
|
|
K::Statistics<float> run(DataSetup setup, int numFile, std::string folder, std::vector<int> gtPath) {
|
|
|
|
std::vector<double> kld_data;
|
|
|
|
// load the floorplan
|
|
Floorplan::IndoorMap* map = Floorplan::Reader::readFromFile(setup.map);
|
|
MyState::map = map;
|
|
|
|
WiFiModelLogDistCeiling WiFiModel(map);
|
|
WiFiModel.loadAPs(map, Settings::WiFiModel::TXP, Settings::WiFiModel::EXP, Settings::WiFiModel::WAF);
|
|
Assert::isFalse(WiFiModel.getAllAPs().empty(), "no AccessPoints stored within the map.xml");
|
|
|
|
//Wi-Fi model new
|
|
// WiFiModelFactory factory(map);
|
|
// WiFiModel* wifimodel= factory.loadXML("/home/toni/Documents/programme/localization/data/wifi/model/eachOptParPos_multimodel.xml");
|
|
// Assert::isFalse(wifimodel->getAllAPs().empty(), "no AccessPoints stored within the map.xml");
|
|
|
|
BeaconModelLogDistCeiling beaconModel(map);
|
|
beaconModel.loadBeaconsFromMap(map, Settings::BeaconModel::TXP, Settings::BeaconModel::EXP, Settings::BeaconModel::WAF);
|
|
//Assert::isFalse(beaconModel.getAllBeacons().empty(), "no Beacons stored within the map.xml");
|
|
|
|
|
|
// build the grid
|
|
std::ifstream inp(setup.grid, std::ifstream::binary);
|
|
Grid<MyNode> grid(20);
|
|
|
|
// grid.dat empty? -> build one and save it
|
|
if (!inp.good() || (inp.peek()&&0) || inp.eof()) {
|
|
std::ofstream onp;
|
|
onp.open(setup.grid);
|
|
GridFactory<MyNode> factory(grid);
|
|
factory.build(map);
|
|
|
|
// add node-importance
|
|
Importance::addImportance(grid);
|
|
|
|
grid.write(onp);
|
|
} else {
|
|
grid.read(inp);
|
|
}
|
|
|
|
// stamp WiFi signal-strengths onto the grid
|
|
WiFiGridEstimator::estimate(grid, WiFiModel, Settings::smartphoneAboveGround);
|
|
|
|
// reading file
|
|
Offline::FileReader fr(setup.training[numFile]);
|
|
|
|
//interpolator for ground truth
|
|
Interpolator<uint64_t, Point3> gtInterpolator = fr.getGroundTruthPath(map, gtPath);
|
|
|
|
//gnuplot plot
|
|
Plotti plot;
|
|
plot.addFloors(map);
|
|
plot.addOutline(map);
|
|
plot.addStairs(map);
|
|
plot.gp << "set autoscale xy\n";
|
|
//plot.addGrid(grid);
|
|
|
|
|
|
// init ctrl and observation
|
|
MyControl ctrl;
|
|
ctrl.resetAfterTransition();
|
|
MyObs obs;
|
|
|
|
//History of all estimated particles. Used for smoothing
|
|
std::vector<std::vector<K::Particle<MyState>>> pfHistory;
|
|
std::vector<int64_t> tsHistory;
|
|
|
|
//filter init
|
|
K::ParticleFilterHistory<MyState, MyControl, MyObs> pf(Settings::numParticles, std::unique_ptr<PFInit>(new PFInit(grid)));
|
|
//K::ParticleFilterHistory<MyState, MyControl, MyObs> pf(Settings::numParticles, std::unique_ptr<PFInitFixed>(new PFInitFixed(grid, GridPoint(1120.0f, 750.0f, 740.0f), 90.0f)));
|
|
|
|
pf.setTransition(std::unique_ptr<PFTrans>(new PFTrans(grid, &ctrl)));
|
|
//pf.setTransition(std::unique_ptr<PFTransKLDSampling>(new PFTransKLDSampling(grid, &ctrl)));
|
|
//pf.setTransition(std::unique_ptr<PFTransSimple>(new PFTransSimple(grid)));
|
|
pf.setEvaluation(std::unique_ptr<PFEval>(new PFEval(WiFiModel, beaconModel, grid)));
|
|
|
|
//resampling
|
|
if(Settings::useKLB){
|
|
pf.setResampling(std::unique_ptr<K::ParticleFilterResamplingDivergence<MyState>>(new K::ParticleFilterResamplingDivergence<MyState>()));
|
|
} else {
|
|
pf.setResampling(std::unique_ptr<K::ParticleFilterResamplingSimple<MyState>>(new K::ParticleFilterResamplingSimple<MyState>()));
|
|
//pf.setResampling(std::unique_ptr<K::ParticleFilterResamplingPercent<MyState>>(new K::ParticleFilterResamplingPercent<MyState>(0.4)));
|
|
//pf.setResampling(std::unique_ptr<NodeResampling<MyState, MyNode>>(new NodeResampling<MyState, MyNode>(*grid)););
|
|
//pf.setResampling(std::unique_ptr<K::ParticleFilterResamplingKLD<MyState>>(new K::ParticleFilterResamplingKLD<MyState>()));
|
|
}
|
|
|
|
pf.setNEffThreshold(0.85);
|
|
|
|
//estimation
|
|
pf.setEstimation(std::unique_ptr<K::ParticleFilterEstimationWeightedAverage<MyState>>(new K::ParticleFilterEstimationWeightedAverage<MyState>()));
|
|
//pf.setEstimation(std::unique_ptr<K::ParticleFilterEstimationRegionalWeightedAverage<MyState>>(new K::ParticleFilterEstimationRegionalWeightedAverage<MyState>()));
|
|
//pf.setEstimation(std::unique_ptr<K::ParticleFilterEstimationOrderedWeightedAverage<MyState>>(new K::ParticleFilterEstimationOrderedWeightedAverage<MyState>(0.5)));
|
|
//pf.setEstimation(std::unique_ptr<K::ParticleFilterEstimationKernelDensity<MyState, 3>>(new K::ParticleFilterEstimationKernelDensity<MyState, 3>()));
|
|
|
|
|
|
/** Smoothing Init */
|
|
K::BackwardSimulation<MyState> bf(Settings::numBSParticles);
|
|
if(Settings::Smoothing::activated){
|
|
|
|
//create the backward smoothing filter
|
|
bf.setSampler( std::unique_ptr<K::CumulativeSampler<MyState>>(new K::CumulativeSampler<MyState>()));
|
|
|
|
bool smoothing_resample = false;
|
|
//bf->setNEffThreshold(1.0);
|
|
if(smoothing_resample)
|
|
bf.setResampling( std::unique_ptr<K::ParticleFilterResamplingSimple<MyState>>(new K::ParticleFilterResamplingSimple<MyState>()) );
|
|
bf.setTransition(std::unique_ptr<BFTrans>( new BFTrans) );
|
|
|
|
//Smoothing estimation
|
|
bf.setEstimation(std::unique_ptr<K::ParticleFilterEstimationWeightedAverage<MyState>>(new K::ParticleFilterEstimationWeightedAverage<MyState>()));
|
|
//bf->setEstimation( std::unique_ptr<K::ParticleFilterEstimationRegionalWeightedAverage<MyState>>(new K::ParticleFilterEstimationRegionalWeightedAverage<MyState>()));
|
|
//bf->setEstimation( std::unique_ptr<K::ParticleFilterEstimationOrderedWeightedAverage<MyState>>(new K::ParticleFilterEstimationOrderedWeightedAverage<MyState>(0.50f)));
|
|
}
|
|
|
|
|
|
|
|
Timestamp lastTimestamp = Timestamp::fromMS(0);
|
|
|
|
StepDetection sd;
|
|
TurnDetection td;
|
|
MotionDetection md;
|
|
ActivityButterPressure act;
|
|
|
|
RelativePressure relBaro;
|
|
relBaro.setCalibrationTimeframe( Timestamp::fromMS(5000) );
|
|
|
|
K::Statistics<float> errorStats;
|
|
K::Statistics<float> errorStatsSmoothing;
|
|
|
|
//file writing for error data
|
|
const long int t = static_cast<long int>(time(NULL));
|
|
const std::string evalDir = errorDir + folder + std::to_string(t);
|
|
if(mkdir(evalDir.c_str(), S_IRWXU | S_IRWXG | S_IROTH | S_IXOTH) == -1){
|
|
Assert::doThrow("Eval folder couldn't be created!");
|
|
}
|
|
|
|
std::ofstream errorFile;
|
|
errorFile.open (evalDir + "/" + std::to_string(numFile) + "_" + std::to_string(t) + ".csv");
|
|
|
|
std::ofstream errorFileSmoothing;
|
|
errorFileSmoothing.open (evalDir + "/" + std::to_string(numFile) + "_" + std::to_string(t) + "_Smoothing.csv");
|
|
|
|
// parse each sensor-value within the offline data
|
|
for (const Offline::Entry& e : fr.getEntries()) {
|
|
|
|
const Timestamp ts = Timestamp::fromMS(e.ts);
|
|
|
|
if (e.type == Offline::Sensor::WIFI) {
|
|
obs.wifi = fr.getWiFiGroupedByTime()[e.idx].data;
|
|
|
|
} else if (e.type == Offline::Sensor::BEACON){
|
|
obs.beacons.entries.push_back(fr.getBeacons()[e.idx].data);
|
|
|
|
// remove to old beacon measurements
|
|
obs.beacons.removeOld(ts);
|
|
|
|
} else if (e.type == Offline::Sensor::ACC) {
|
|
if (sd.add(ts, fr.getAccelerometer()[e.idx].data)) {
|
|
++ctrl.numStepsSinceLastTransition;
|
|
}
|
|
const Offline::TS<AccelerometerData>& _acc = fr.getAccelerometer()[e.idx];
|
|
td.addAccelerometer(ts, _acc.data);
|
|
|
|
} else if (e.type == Offline::Sensor::GYRO) {
|
|
const Offline::TS<GyroscopeData>& _gyr = fr.getGyroscope()[e.idx];
|
|
const float delta_gyro = td.addGyroscope(ts, _gyr.data);
|
|
|
|
ctrl.turnSinceLastTransition_rad += delta_gyro;
|
|
|
|
} else if (e.type == Offline::Sensor::BARO) {
|
|
relBaro.add(ts, fr.getBarometer()[e.idx].data);
|
|
obs.relativePressure = relBaro.getPressureRealtiveToStart();
|
|
obs.sigmaPressure = relBaro.getSigma();
|
|
|
|
//activity recognition
|
|
obs.activity = act.add(ts, fr.getBarometer()[e.idx].data);
|
|
//activity for transition
|
|
|
|
} else if (e.type == Offline::Sensor::LIN_ACC) {
|
|
md.addLinearAcceleration(ts, fr.getLinearAcceleration()[e.idx].data);
|
|
|
|
} else if (e.type == Offline::Sensor::GRAVITY) {
|
|
md.addGravity(ts, fr.getGravity()[e.idx].data);
|
|
Eigen::Vector2f curVec = md.getCurrentMotionAxis();
|
|
ctrl.motionDeltaAngle_rad = md.getMotionChangeInRad();
|
|
}
|
|
|
|
if (ts.ms() - lastTimestamp.ms() > 500) {
|
|
|
|
|
|
/** filtering stuff */
|
|
obs.currentTime = ts;
|
|
MyState est = pf.update(&ctrl, obs);
|
|
|
|
Point3 estPos = est.position.inMeter();
|
|
Point3 gtPos = gtInterpolator.get(static_cast<uint64_t>(ts.ms()));
|
|
|
|
/** plotting stuff */
|
|
plot.pInterest.clear();
|
|
|
|
plot.setEst(estPos);
|
|
plot.setGT(gtPos);
|
|
plot.addEstimationNode(estPos);
|
|
plot.addParticles(pf.getParticles());
|
|
|
|
/** error calculation stuff */
|
|
float err_m = gtPos.getDistance(estPos);
|
|
errorStats.add(err_m);
|
|
errorFile << err_m << "\n";
|
|
|
|
|
|
/** smoothing stuff */
|
|
if(Settings::Smoothing::activated){
|
|
//save the current estimation for later smoothing.
|
|
pfHistory.push_back(pf.getNonResamplingParticles());
|
|
tsHistory.push_back(ts.ms());
|
|
|
|
//backward filtering
|
|
MyState estBF = est;
|
|
if(pfHistory.size() > Settings::Smoothing::lag){
|
|
|
|
bf.reset();
|
|
|
|
//use every n-th (std = 1) particle set of the history within a given lag (std = 5)
|
|
for(int i = 0; i <= Settings::Smoothing::lag; ++i){
|
|
|
|
((BFTrans*)bf.getTransition())->setCurrentTime(tsHistory[(tsHistory.size() - 1) - i]);
|
|
estBF = bf.update(pfHistory[(pfHistory.size() - 1) - i]);
|
|
}
|
|
}
|
|
|
|
Point3 estPosSmoothing = estBF.position.inMeter();
|
|
Point3 gtPosSmoothed = gtInterpolator.get(static_cast<uint64_t>(tsHistory[(tsHistory.size() - 1) - Settings::Smoothing::lag]));
|
|
|
|
//plot
|
|
plot.addEstimationNodeSmoothed(estPosSmoothing);
|
|
|
|
// error between GT and smoothing
|
|
float errSmoothing_m = gtPosSmoothed.getDistance(estPosSmoothing);
|
|
errorStatsSmoothing.add(errSmoothing_m);
|
|
errorFileSmoothing << errSmoothing_m << "\n";
|
|
}
|
|
|
|
//plot misc
|
|
plot.setTimeInMinute(static_cast<int>(ts.sec()) / 60, static_cast<int>(static_cast<int>(ts.sec())%60));
|
|
|
|
if(Settings::useKLB){
|
|
plot.gp << "set label 1002 at screen 0.04, 0.94 'KLD: " << ":" << kld_data.back() << "'\n";
|
|
}
|
|
plot.gp << "set label 1002 at screen 0.98, 0.98 'act:" << obs.activity << "'\n";
|
|
|
|
/** Draw everything */
|
|
plot.show();
|
|
usleep(10*10);
|
|
|
|
lastTimestamp = ts;
|
|
|
|
// reset control
|
|
ctrl.resetAfterTransition();
|
|
}
|
|
|
|
}
|
|
|
|
errorFile.close();
|
|
|
|
std::cout << "Statistical Analysis Filtering: " << std::endl;
|
|
std::cout << "Median: " << errorStats.getMedian() << " Average: " << errorStats.getAvg() << " Std: " << errorStats.getStdDev() << std::endl;
|
|
|
|
std::cout << "Statistical Analysis Smoothing: " << std::endl;
|
|
std::cout << "Median: " << errorStatsSmoothing.getMedian() << " Average: " << errorStatsSmoothing.getAvg() << " Std: " << errorStatsSmoothing.getStdDev() << std::endl;
|
|
|
|
//Write the current plotti buffer into file
|
|
std::ofstream plotFile;
|
|
plotFile.open(evalDir + "/plot_" + std::to_string(numFile) + "_" + std::to_string(t) + ".gp");
|
|
plot.saveToFile(plotFile);
|
|
plotFile.close();
|
|
|
|
for(int i = 0; i < map->floors.size(); ++i){
|
|
plot.printSingleFloor(evalDir + "/image" + std::to_string(numFile) + "_" + std::to_string(t), i);
|
|
plot.show();
|
|
usleep(10*10);
|
|
}
|
|
|
|
plot.printSideView(evalDir + "/image" + std::to_string(numFile) + "_" + std::to_string(t), 90);
|
|
plot.show();
|
|
|
|
plot.printSideView(evalDir + "/image" + std::to_string(numFile) + "_" + std::to_string(t), 0);
|
|
plot.show();
|
|
|
|
plot.printOverview(evalDir + "/image" + std::to_string(numFile) + "_" + std::to_string(t));
|
|
plot.show();
|
|
|
|
|
|
/** Draw KLB */
|
|
K::Gnuplot gp;
|
|
K::GnuplotPlot plotkld;
|
|
K::GnuplotPlotElementLines lines;
|
|
if(Settings::useKLB){
|
|
|
|
std::string path = evalDir + "/image" + std::to_string(numFile) + "_" + std::to_string(t);
|
|
gp << "set terminal png size 1280,720\n";
|
|
gp << "set output '" << path << "_shennendistance.png'\n";
|
|
|
|
for(int i=0; i < kld_data.size()-1; ++i){
|
|
|
|
K::GnuplotPoint2 p1(i, kld_data[i]);
|
|
K::GnuplotPoint2 p2(i+1, kld_data[i+1]);
|
|
|
|
lines.addSegment(p1, p2);
|
|
}
|
|
|
|
plotkld.add(&lines);
|
|
gp.draw(plotkld);
|
|
gp.flush();
|
|
}
|
|
|
|
std::cout << "finished" << std::endl;
|
|
sleep(1);
|
|
|
|
return errorStats;
|
|
|
|
}
|
|
|
|
int main(int argc, char** argv) {
|
|
|
|
K::Statistics<float> statsAVG;
|
|
K::Statistics<float> statsMedian;
|
|
K::Statistics<float> statsSTD;
|
|
K::Statistics<float> statsQuantil;
|
|
K::Statistics<float> tmp;
|
|
|
|
for(int i = 0; i < 1; ++i){
|
|
|
|
tmp = run(data.SecondFloorOnly, 2, "Wifi-Dongle-Test", Settings::Path_DongleTest::path3);
|
|
statsMedian.add(tmp.getMedian());
|
|
statsAVG.add(tmp.getAvg());
|
|
statsSTD.add(tmp.getStdDev());
|
|
statsQuantil.add(tmp.getQuantile(0.75));
|
|
|
|
std::cout << "Iteration " << i << " completed" << std::endl;;
|
|
}
|
|
|
|
std::cout << "==========================================================" << std::endl;
|
|
std::cout << "Average of all statistical data: " << std::endl;
|
|
std::cout << "Median: " << statsMedian.getAvg() << std::endl;
|
|
std::cout << "Average: " << statsAVG.getAvg() << std::endl;
|
|
std::cout << "Standard Deviation: " << statsSTD.getAvg() << std::endl;
|
|
std::cout << "75 Quantil: " << statsQuantil.getAvg() << std::endl;
|
|
std::cout << "==========================================================" << std::endl;
|
|
|
|
//EDIT THIS EDIT THIS EDIT THIS EDIT THIS EDIT THIS EDIT THIS EDIT THIS EDIT THIS
|
|
std::ofstream finalStatisticFile;
|
|
finalStatisticFile.open (errorDir + "/tmp.csv");
|
|
|
|
finalStatisticFile << "Average of all statistical data: \n";
|
|
finalStatisticFile << "Median: " << statsMedian.getAvg() << "\n";
|
|
finalStatisticFile << "Average: " << statsAVG.getAvg() << "\n";
|
|
finalStatisticFile << "Standard Deviation: " << statsSTD.getAvg() << "\n";
|
|
finalStatisticFile << "75 Quantil: " << statsQuantil.getAvg() << "\n";
|
|
|
|
finalStatisticFile.close();
|
|
//EDIT THIS EDIT THIS EDIT THIS EDIT THIS EDIT THIS EDIT THIS EDIT THIS EDIT THIS
|
|
|
|
}
|