Template Class AbsoluteConstraint

Inheritance Relationships

Base Type

  • public fuse_core::Constraint

Class Documentation

template<class Variable>
class AbsoluteConstraint : public fuse_core::Constraint

A constraint that represents prior information about a variable or a direct measurement of the variable.

This type of constraint arises in many situations. In mapping it is common to define the very first pose as the origin. Some sensors, such as an IMU, provide direct measurements of certain state variables (e.g. linear acceleration). And localization systems often match laserscans to a prior map (scan-to-map measurements). This constraint holds the measured variable value and the measurement uncertainty/covariance.

Public Functions

AbsoluteConstraint() = default

Default constructor.

AbsoluteConstraint(const std::string &source, const Variable &variable, const fuse_core::VectorXd &mean, const fuse_core::MatrixXd &covariance)

Create a constraint using a measurement/prior of all dimensions of the target variable.

Parameters:
  • source[in] The name of the sensor or motion model that generated this constraint

  • variable[in] An object derived from fuse_core::Variable.

  • mean[in] The measured/prior value of all variable dimensions

  • covariance[in] The measurement/prior uncertainty of all variable dimensions

AbsoluteConstraint(const std::string &source, const Variable &variable, const fuse_core::VectorXd &partial_mean, const fuse_core::MatrixXd &partial_covariance, const std::vector<size_t> &indices)

Create a constraint using a measurement/prior of only a partial set of dimensions of the target variable.

Parameters:
  • source[in] The name of the sensor or motion model that generated this constraint

  • variable[in] An object derived from fuse_core::Variable.

  • partial_mean[in] The measured value of the subset of dimensions in the order defined by indices

  • partial_covariance[in] The uncertainty of the subset of dimensions in the order defined by indices.

  • indices[in] The set of indices corresponding to the measured dimensions

virtual ~AbsoluteConstraint() = default

Destructor.

inline const fuse_core::VectorXd &mean() const

Read-only access to the measured/prior vector of mean values.

All dimensions are present, even if only a partial set of dimensions were measured. Dimensions are in the order defined by the variable, not the order defined by the indices parameter. All unmeasured variable dimensions are set to zero.

inline const fuse_core::MatrixXd &sqrtInformation() const

Read-only access to the square root information matrix.

Dimensions are in the order defined by the variable, not the order defined by the indices parameter. The square root information matrix will have size measured_dimensions X variable_dimensions. If only a partial set of dimensions are measured, then this matrix will not be square.

fuse_core::MatrixXd covariance() const

Compute the measurement covariance matrix.

Dimensions are in the order defined by the variable, not the order defined by the indices parameter. The covariance matrix will always be square, with size variable_dimensions X variable_dimensions. If only a subset of dimensions are measured, then some rows/columns will be zero. This will result in a rank-deficient covariance matrix. You have been warned.

void print(std::ostream &stream = std::cout) const override

Print a human-readable description of the constraint to the provided stream.

Parameters:

stream[out] The stream to write to. Defaults to stdout.

ceres::CostFunction *costFunction() const override

Construct an instance of this constraint’s cost function.

The function caller will own the new cost function instance. It is the responsibility of the caller to delete the cost function object when it is no longer needed. If the pointer is provided to a Ceres::Problem object, the Ceres::Problem object will takes ownership of the pointer and delete it during destruction.

Returns:

A base pointer to an instance of a derived CostFunction.

inline ceres::CostFunction *costFunction() const

Protected Attributes

fuse_core::VectorXd mean_

The measured/prior mean vector for this variable.

fuse_core::MatrixXd sqrt_information_

The square root information matrix.

Friends

friend class boost::serialization::access