gtsam/matlab/unstable_examples/IMUKittiExampleAdvanced.m

close all
clc

import gtsam.*;
disp('Example of application of ISAM2 for visual-inertial navigation on the KITTI VISION BENCHMARK SUITE (http://www.computervisiononline.com/dataset/kitti-vision-benchmark-suite)')

%% Read metadata and compute relative sensor pose transforms
% IMU metadata
disp('-- Reading sensor metadata')
IMU_metadata = importdata(findExampleDataFile('KittiEquivBiasedImu_metadata.txt'));
IMU_metadata = cell2struct(num2cell(IMU_metadata.data), IMU_metadata.colheaders, 2);
IMUinBody = Pose3.Expmap([IMU_metadata.BodyPtx; IMU_metadata.BodyPty; IMU_metadata.BodyPtz;
  IMU_metadata.BodyPrx; IMU_metadata.BodyPry; IMU_metadata.BodyPrz; ]);
if ~IMUinBody.equals(Pose3, 1e-5)
  error 'Currently only support IMUinBody is identity, i.e. IMU and body frame are the same';
end

% VO metadata
VO_metadata = importdata(findExampleDataFile('KittiRelativePose_metadata.txt'));
VO_metadata = cell2struct(num2cell(VO_metadata.data), VO_metadata.colheaders, 2);
VOinBody = Pose3.Expmap([VO_metadata.BodyPtx; VO_metadata.BodyPty; VO_metadata.BodyPtz;
  VO_metadata.BodyPrx; VO_metadata.BodyPry; VO_metadata.BodyPrz; ]);
VOinIMU = IMUinBody.inverse().compose(VOinBody);

% GPS metadata
GPS_metadata = importdata(findExampleDataFile('KittiGps_metadata.txt'));
GPS_metadata = cell2struct(num2cell(GPS_metadata.data), GPS_metadata.colheaders, 2);
GPSinBody = Pose3.Expmap([GPS_metadata.BodyPtx; GPS_metadata.BodyPty; GPS_metadata.BodyPtz;
  GPS_metadata.BodyPrx; GPS_metadata.BodyPry; GPS_metadata.BodyPrz; ]);
GPSinIMU = IMUinBody.inverse().compose(GPSinBody);

%% Read data and change coordinate frame of GPS and VO measurements to IMU frame
disp('-- Reading sensor data from file')
% IMU data
IMU_data = importdata(findExampleDataFile('KittiEquivBiasedImu.txt'));
IMU_data = cell2struct(num2cell(IMU_data.data), IMU_data.colheaders, 2);
imum = cellfun(@(x) x', num2cell([ [IMU_data.accelX]' [IMU_data.accelY]' [IMU_data.accelZ]' [IMU_data.omegaX]' [IMU_data.omegaY]' [IMU_data.omegaZ]' ], 2), 'UniformOutput', false);
[IMU_data.acc_omega] = deal(imum{:});
%IMU_data = rmfield(IMU_data, { 'accelX' 'accelY' 'accelZ' 'omegaX' 'omegaY' 'omegaZ' });
clear imum

% VO data
VO_data = importdata(findExampleDataFile('KittiRelativePose.txt'));
VO_data = cell2struct(num2cell(VO_data.data), VO_data.colheaders, 2);
% Merge relative pose fields and convert to Pose3
logposes = [ [VO_data.dtx]' [VO_data.dty]' [VO_data.dtz]' [VO_data.drx]' [VO_data.dry]' [VO_data.drz]' ];
logposes = num2cell(logposes, 2);
relposes = arrayfun(@(x) {gtsam.Pose3.Expmap(x{:}')}, logposes);
relposes = arrayfun(@(x) {VOinIMU.compose(x{:}).compose(VOinIMU.inverse())}, relposes);
[VO_data.RelativePose] = deal(relposes{:});
VO_data = rmfield(VO_data, { 'dtx' 'dty' 'dtz' 'drx' 'dry' 'drz' });
noiseModelVO = noiseModel.Diagonal.Sigmas([ VO_metadata.RotationSigma * [1;1;1]; VO_metadata.TranslationSigma * [1;1;1] ]);
clear logposes relposes

% % % GPS data
% % GPS_data = importdata('KittiGps.txt');
% % GPS_data = cell2struct(num2cell(GPS_data.data), GPS_data.colheaders, 2);
% % % Convert GPS from lat/long to meters
% % [ x, y, ~ ] = deg2utm( [GPS_data.Latitude], [GPS_data.Longitude] );
% % for i = 1:numel(x)
% %   GPS_data(i).Position = gtsam.Point3(x(i), y(i), GPS_data(i).Altitude);
% % end
% % % % Calculate GPS sigma in meters
% % % [ xSig, ySig, ~ ] = deg2utm( [GPS_data.Latitude] + [GPS_data.PositionSigma], ...
% % %     [GPS_data.Longitude] + [GPS_data.PositionSigma]);
% % % xSig = xSig - x;
% % % ySig = ySig - y;
% % %% Start at time of first GPS measurement
% % % firstGPSPose = 1;

%% Get initial conditions for the estimated trajectory
% % % currentPoseGlobal = Pose3(Rot3, GPS_data(firstGPSPose).Position); % initial pose is the reference frame (navigation frame)
currentPoseGlobal = Pose3;
currentVelocityGlobal = [0;0;0]; % the vehicle is stationary at the beginning
currentBias = imuBias.ConstantBias(zeros(3,1), zeros(3,1));
sigma_init_x = noiseModel.Isotropic.Sigma(6, 0.01);
sigma_init_v = noiseModel.Isotropic.Sigma(3, 1000.0);
sigma_init_b = noiseModel.Isotropic.Sigma(6, 0.01);
g = [0;0;-9.8];
w_coriolis = [0;0;0];

%% Solver object
isamParams = ISAM2Params;
isamParams.setFactorization('QR');
isamParams.relinearizeSkip = 1;
isam = gtsam.ISAM2(isamParams);
newFactors = NonlinearFactorGraph;
newValues = Values;

%% Main loop:
% (1) we read the measurements
% (2) we create the corresponding factors in the graph
% (3) we solve the graph to obtain and optimal estimate of robot trajectory
timestamps = sortrows( [ ...
  [VO_data.Time]' 1*ones(length([VO_data.Time]), 1); ...
% %   %[GPS_data.Time]' 2*ones(length([GPS_data.Time]), 1); ...
  ], 1); % this are the time-stamps at which we want to initialize a new node in the graph

timestamps = timestamps(15:end,:); % there seem to be issues with the initial IMU measurements
IMUtimes = [IMU_data.Time];
VOPoseKeys = []; % here we store the keys of the poses involved in VO (between) factors

for measurementIndex = 1:length(timestamps)
  
  % At each non=IMU measurement we initialize a new node in the graph
  currentPoseKey = symbol('x',measurementIndex);
  currentVelKey =  symbol('v',measurementIndex);
  currentBiasKey = symbol('b',measurementIndex);
  t = timestamps(measurementIndex, 1);
  type = timestamps(measurementIndex, 2);
  
  %% bookkeeping
  if type == 1 % we store the keys corresponding to VO measurements
    VOPoseKeys = [VOPoseKeys; currentPoseKey];
  end
  
  if measurementIndex == 1
    %% Create initial estimate and prior on initial pose, velocity, and biases
    newValues.insert(currentPoseKey, currentPoseGlobal);
    newValues.insert(currentVelKey, currentVelocityGlobal);
    newValues.insert(currentBiasKey, currentBias);
    newFactors.add(PriorFactorPose3(currentPoseKey, currentPoseGlobal, sigma_init_x));
    newFactors.add(PriorFactorVector(currentVelKey, currentVelocityGlobal, sigma_init_v));
    newFactors.add(PriorFactorConstantBias(currentBiasKey, currentBias, sigma_init_b));
  else
    t_previous = timestamps(measurementIndex-1, 1);
    %% Summarize IMU data between the previous GPS measurement and now
    IMUindices = find(IMUtimes >= t_previous & IMUtimes <= t);
    
    if ~isempty(IMUindices) % if there are IMU measurements to integrate
      currentSummarizedMeasurement = gtsam.ImuFactorPreintegratedMeasurements( ...
        currentBias, IMU_metadata.AccelerometerSigma.^2 * eye(3), ...
        IMU_metadata.GyroscopeSigma.^2 * eye(3), IMU_metadata.IntegrationSigma.^2 * eye(3));

      for imuIndex = IMUindices
        accMeas = [ IMU_data(imuIndex).accelX; IMU_data(imuIndex).accelY; IMU_data(imuIndex).accelZ ];
        omegaMeas = [ IMU_data(imuIndex).omegaX; IMU_data(imuIndex).omegaY; IMU_data(imuIndex).omegaZ ];
        deltaT = IMU_data(imuIndex).dt;
        currentSummarizedMeasurement.integrateMeasurement(accMeas, omegaMeas, deltaT);
      end
      
      % Create IMU factor
      newFactors.add(ImuFactor( ...
        currentPoseKey-1, currentVelKey-1, ...
        currentPoseKey, currentVelKey, ...
        currentBiasKey, currentSummarizedMeasurement, g, w_coriolis));
      
    else % if there are no IMU measurements
      error('no IMU measurements in [t_previous, t]')
    end
    
    % LC: sigma_init_b is wrong: this should be some uncertainty on bias evolution given in the IMU metadata
    newFactors.add(BetweenFactorConstantBias(currentBiasKey-1, currentBiasKey, imuBias.ConstantBias(zeros(3,1), zeros(3,1)), sigma_init_b));
    
    %% Create GPS factor
    if type == 2
      newFactors.add(PriorFactorPose3(currentPoseKey, Pose3(currentPoseGlobal.rotation, GPS_data(measurementIndex).Position), ...
        noiseModel.Diagonal.Precisions([ zeros(3,1); 1./(GPS_data(measurementIndex).PositionSigma).^2*ones(3,1) ])));
    end
    
    %% Create VO factor
    if type == 1
      VOpose = VO_data(measurementIndex).RelativePose;
      newFactors.add(BetweenFactorPose3(VOPoseKeys(end-1), VOPoseKeys(end), VOpose, noiseModelVO));
    end
    
    % Add initial value
    % newValues.insert(currentPoseKey, Pose3(currentPoseGlobal.rotation, GPS_data(measurementIndex).Position));
    newValues.insert(currentPoseKey,currentPoseGlobal);
    newValues.insert(currentVelKey, currentVelocityGlobal);
    newValues.insert(currentBiasKey, currentBias);
    
    % Update solver
    % =======================================================================
    isam.update(newFactors, newValues);
    newFactors = NonlinearFactorGraph;
    newValues = Values;
    
    if rem(measurementIndex,100)==0 % plot every 100 time steps
      cla;
      plot3DTrajectory(isam.calculateEstimate, 'g-');
      axis equal
      drawnow;
    end
    % ======================================================================= 
    currentPoseGlobal = isam.calculateEstimate(currentPoseKey);
    currentVelocityGlobal = isam.calculateEstimate(currentVelKey);
    currentBias = isam.calculateEstimate(currentBiasKey);   
  end
   
end % end main loop