| 1 | /* ---------------------------------------------------------------------------- |
| 2 | |
| 3 | * GTSAM Copyright 2010, Georgia Tech Research Corporation, |
| 4 | * Atlanta, Georgia 30332-0415 |
| 5 | * All Rights Reserved |
| 6 | * Authors: Frank Dellaert, et al. (see THANKS for the full author list) |
| 7 | |
| 8 | * See LICENSE for the license information |
| 9 | |
| 10 | * -------------------------------------------------------------------------- */ |
| 11 | |
| 12 | /** |
| 13 | * @file Pose2SLAMExampleExpressions.cpp |
| 14 | * @brief Expressions version of Pose2SLAMExample.cpp |
| 15 | * @date Oct 2, 2014 |
| 16 | * @author Frank Dellaert |
| 17 | */ |
| 18 | |
| 19 | // The two new headers that allow using our Automatic Differentiation Expression framework |
| 20 | #include <gtsam/slam/expressions.h> |
| 21 | #include <gtsam/nonlinear/ExpressionFactorGraph.h> |
| 22 | |
| 23 | // For an explanation of headers below, please see Pose2SLAMExample.cpp |
| 24 | #include <gtsam/slam/BetweenFactor.h> |
| 25 | #include <gtsam/geometry/Pose2.h> |
| 26 | #include <gtsam/nonlinear/GaussNewtonOptimizer.h> |
| 27 | #include <gtsam/nonlinear/Marginals.h> |
| 28 | |
| 29 | using namespace std; |
| 30 | using namespace gtsam; |
| 31 | |
| 32 | int main(int argc, char** argv) { |
| 33 | // 1. Create a factor graph container and add factors to it |
| 34 | ExpressionFactorGraph graph; |
| 35 | |
| 36 | // Create Expressions for unknowns |
| 37 | Pose2_ x1(1), x2(2), x3(3), x4(4), x5(5); |
| 38 | |
| 39 | // 2a. Add a prior on the first pose, setting it to the origin |
| 40 | auto priorNoise = noiseModel::Diagonal::Sigmas(sigmas: Vector3(0.3, 0.3, 0.1)); |
| 41 | graph.addExpressionFactor(h: x1, z: Pose2(0, 0, 0), R: priorNoise); |
| 42 | |
| 43 | // For simplicity, we use the same noise model for odometry and loop closures |
| 44 | auto model = noiseModel::Diagonal::Sigmas(sigmas: Vector3(0.2, 0.2, 0.1)); |
| 45 | |
| 46 | // 2b. Add odometry factors |
| 47 | graph.addExpressionFactor(h: between(t1: x1, t2: x2), z: Pose2(2, 0, 0), R: model); |
| 48 | graph.addExpressionFactor(h: between(t1: x2, t2: x3), z: Pose2(2, 0, M_PI_2), R: model); |
| 49 | graph.addExpressionFactor(h: between(t1: x3, t2: x4), z: Pose2(2, 0, M_PI_2), R: model); |
| 50 | graph.addExpressionFactor(h: between(t1: x4, t2: x5), z: Pose2(2, 0, M_PI_2), R: model); |
| 51 | |
| 52 | // 2c. Add the loop closure constraint |
| 53 | graph.addExpressionFactor(h: between(t1: x5, t2: x2), z: Pose2(2, 0, M_PI_2), R: model); |
| 54 | graph.print(str: "\nFactor Graph:\n" ); // print |
| 55 | |
| 56 | // 3. Create the data structure to hold the initialEstimate estimate to the |
| 57 | // solution For illustrative purposes, these have been deliberately set to |
| 58 | // incorrect values |
| 59 | Values initialEstimate; |
| 60 | initialEstimate.insert(j: 1, val: Pose2(0.5, 0.0, 0.2)); |
| 61 | initialEstimate.insert(j: 2, val: Pose2(2.3, 0.1, -0.2)); |
| 62 | initialEstimate.insert(j: 3, val: Pose2(4.1, 0.1, M_PI_2)); |
| 63 | initialEstimate.insert(j: 4, val: Pose2(4.0, 2.0, M_PI)); |
| 64 | initialEstimate.insert(j: 5, val: Pose2(2.1, 2.1, -M_PI_2)); |
| 65 | initialEstimate.print(str: "\nInitial Estimate:\n" ); // print |
| 66 | |
| 67 | // 4. Optimize the initial values using a Gauss-Newton nonlinear optimizer |
| 68 | GaussNewtonParams parameters; |
| 69 | parameters.relativeErrorTol = 1e-5; |
| 70 | parameters.maxIterations = 100; |
| 71 | GaussNewtonOptimizer optimizer(graph, initialEstimate, parameters); |
| 72 | Values result = optimizer.optimize(); |
| 73 | result.print(str: "Final Result:\n" ); |
| 74 | |
| 75 | // 5. Calculate and print marginal covariances for all variables |
| 76 | cout.precision(prec: 3); |
| 77 | Marginals marginals(graph, result); |
| 78 | cout << "x1 covariance:\n" << marginals.marginalCovariance(variable: 1) << endl; |
| 79 | cout << "x2 covariance:\n" << marginals.marginalCovariance(variable: 2) << endl; |
| 80 | cout << "x3 covariance:\n" << marginals.marginalCovariance(variable: 3) << endl; |
| 81 | cout << "x4 covariance:\n" << marginals.marginalCovariance(variable: 4) << endl; |
| 82 | cout << "x5 covariance:\n" << marginals.marginalCovariance(variable: 5) << endl; |
| 83 | |
| 84 | return 0; |
| 85 | } |
| 86 | |