GTSAM
4.0.2
C++ library for smoothing and mapping (SAM)
|
#include <SmartStereoProjectionFactor.h>
Public Types | |
typedef std::shared_ptr< SmartStereoProjectionFactor > | shared_ptr |
shorthand for a smart pointer to a factor | |
typedef CameraSet< StereoCamera > | Cameras |
Vector of cameras. | |
typedef PinholeCamera< Cal3_S2 > | MonoCamera |
Vector of monocular cameras (stereo treated as 2 monocular) | |
typedef CameraSet< MonoCamera > | MonoCameras |
typedef MonoCamera::MeasurementVector | MonoMeasurements |
typedef Eigen::Matrix< double, ZDim, Dim > | MatrixZD |
typedef std::vector< MatrixZD, Eigen::aligned_allocator< MatrixZD > > | FBlocks |
typedef KeyVector::iterator | iterator |
Iterator over keys. | |
typedef KeyVector::const_iterator | const_iterator |
Const iterator over keys. | |
Public Member Functions | |
SmartStereoProjectionFactor (const SharedNoiseModel &sharedNoiseModel, const SmartStereoProjectionParams ¶ms=SmartStereoProjectionParams(), const std::optional< Pose3 > body_P_sensor={}) | |
~SmartStereoProjectionFactor () override | |
void | print (const std::string &s="", const KeyFormatter &keyFormatter=DefaultKeyFormatter) const override |
bool | equals (const NonlinearFactor &p, double tol=1e-9) const override |
equals | |
bool | decideIfTriangulate (const Cameras &cameras) const |
Check if the new linearization point_ is the same as the one used for previous triangulation. | |
TriangulationResult | triangulateSafe (const Cameras &cameras) const |
triangulateSafe | |
bool | triangulateForLinearize (const Cameras &cameras) const |
triangulate | |
std::shared_ptr< RegularHessianFactor< Base::Dim > > | createHessianFactor (const Cameras &cameras, const double lambda=0.0, bool diagonalDamping=false) const |
linearize returns a Hessianfactor that is an approximation of error(p) | |
std::shared_ptr< JacobianFactor > | createJacobianSVDFactor (const Cameras &cameras, double lambda) const |
different (faster) way to compute Jacobian factor | |
std::shared_ptr< GaussianFactor > | linearizeDamped (const Cameras &cameras, const double lambda=0.0) const |
std::shared_ptr< GaussianFactor > | linearizeDamped (const Values &values, const double lambda=0.0) const |
std::shared_ptr< GaussianFactor > | linearize (const Values &values) const override |
linearize | |
bool | triangulateAndComputeE (Matrix &E, const Cameras &cameras) const |
bool | triangulateAndComputeE (Matrix &E, const Values &values) const |
void | computeJacobiansWithTriangulatedPoint (FBlocks &Fs, Matrix &E, Vector &b, const Cameras &cameras) const |
bool | triangulateAndComputeJacobians (FBlocks &Fs, Matrix &E, Vector &b, const Values &values) const |
Version that takes values, and creates the point. | |
bool | triangulateAndComputeJacobiansSVD (FBlocks &Fs, Matrix &Enull, Vector &b, const Values &values) const |
takes values | |
Vector | reprojectionErrorAfterTriangulation (const Values &values) const |
Calculate vector of re-projection errors, before applying noise model. | |
double | totalReprojectionError (const Cameras &cameras, std::optional< Point3 > externalPoint={}) const |
double | error (const Values &values) const override |
Calculate total reprojection error. | |
void | correctForMissingMeasurements (const Cameras &cameras, Vector &ue, typename Cameras::FBlocks *Fs=nullptr, Matrix *E=nullptr) const override |
TriangulationResult | point () const |
TriangulationResult | point (const Values &values) const |
bool | isValid () const |
Is result valid? | |
bool | isDegenerate () const |
bool | isPointBehindCamera () const |
bool | isOutlier () const |
bool | isFarPoint () const |
void | add (const Z &measured, const Key &key) |
void | add (const ZVector &measurements, const KeyVector &cameraKeys) |
Add a bunch of measurements, together with the camera keys. | |
void | add (const SFM_TRACK &trackToAdd) |
size_t | dim () const override |
Return the dimension (number of rows!) of the factor. | |
const ZVector & | measured () const |
Return the 2D measurements (ZDim, in general). | |
virtual Cameras | cameras (const Values &values) const |
Collect all cameras: important that in key order. | |
Vector | unwhitenedError (const Cameras &cameras, const POINT &point, typename Cameras::FBlocks *Fs=nullptr, Matrix *E=nullptr) const |
Vector | unwhitenedError (const Cameras &cameras, const POINT &point, OptArgs &&... optArgs) const |
void | correctForMissingMeasurements (const Cameras &cameras, Vector &ue, OptArgs &&... optArgs) const |
Vector | whitenedError (const Cameras &cameras, const POINT &point) const |
double | totalReprojectionError (const Cameras &cameras, const POINT &point) const |
void | computeJacobians (FBlocks &Fs, Matrix &E, Vector &b, const Cameras &cameras, const POINT &point) const |
void | computeJacobiansSVD (FBlocks &Fs, Matrix &Enull, Vector &b, const Cameras &cameras, const POINT &point) const |
std::shared_ptr< RegularHessianFactor< Dim > > | createHessianFactor (const Cameras &cameras, const Point3 &point, const double lambda=0.0, bool diagonalDamping=false) const |
Linearize to a Hessianfactor. | |
void | updateAugmentedHessian (const Cameras &cameras, const Point3 &point, const double lambda, bool diagonalDamping, SymmetricBlockMatrix &augmentedHessian, const KeyVector allKeys) const |
void | whitenJacobians (FBlocks &F, Matrix &E, Vector &b) const |
Whiten the Jacobians computed by computeJacobians using noiseModel_. | |
std::shared_ptr< RegularImplicitSchurFactor< StereoCamera > > | createRegularImplicitSchurFactor (const Cameras &cameras, const Point3 &point, double lambda=0.0, bool diagonalDamping=false) const |
Return Jacobians as RegularImplicitSchurFactor with raw access. | |
std::shared_ptr< JacobianFactorQ< Dim, ZDim > > | createJacobianQFactor (const Cameras &cameras, const Point3 &point, double lambda=0.0, bool diagonalDamping=false) const |
Return Jacobians as JacobianFactorQ. | |
std::shared_ptr< JacobianFactor > | createJacobianSVDFactor (const Cameras &cameras, const Point3 &point, double lambda=0.0) const |
Pose3 | body_P_sensor () const |
Testable | |
bool | equals (const This &other, double tol=1e-9) const |
check equality | |
virtual void | printKeys (const std::string &s="Factor", const KeyFormatter &formatter=DefaultKeyFormatter) const |
print only keys | |
Standard Interface | |
double | error (const HybridValues &c) const override |
virtual bool | active (const Values &) const |
virtual shared_ptr | clone () const |
virtual shared_ptr | rekey (const std::map< Key, Key > &rekey_mapping) const |
virtual shared_ptr | rekey (const KeyVector &new_keys) const |
virtual bool | sendable () const |
Standard Interface | |
bool | empty () const |
Whether the factor is empty (involves zero variables). | |
Key | front () const |
First key. | |
Key | back () const |
Last key. | |
const_iterator | find (Key key) const |
find | |
const KeyVector & | keys () const |
Access the factor's involved variable keys. | |
const_iterator | begin () const |
const_iterator | end () const |
size_t | size () const |
Advanced Interface | |
KeyVector & | keys () |
iterator | begin () |
iterator | end () |
Static Public Member Functions | |
static Matrix | PointCov (const Matrix &E) |
Computes Point Covariance P from the "point Jacobian" E. | |
static void | FillDiagonalF (const FBlocks &Fs, Matrix &F) |
Create BIG block-diagonal matrix F from Fblocks. | |
Static Public Attributes | |
static const int | Dim |
Camera dimension. | |
static const int | ZDim |
Measurement dimension. | |
Static Protected Member Functions | |
Standard Constructors | |
template<typename CONTAINER > | |
static Factor | FromKeys (const CONTAINER &keys) |
template<typename ITERATOR > | |
static Factor | FromIterators (ITERATOR first, ITERATOR last) |
Protected Attributes | |
SharedIsotropic | noiseModel_ |
ZVector | measured_ |
std::optional< Pose3 > | body_P_sensor_ |
Pose of the camera in the body frame. | |
FBlocks | Fs |
KeyVector | keys_ |
The keys involved in this factor. | |
Parameters | |
const SmartStereoProjectionParams | params_ |
Caching triangulation | |
TriangulationResult | result_ |
result from triangulateSafe | |
std::vector< Pose3 > | cameraPosesTriangulation_ |
current triangulation poses | |
SmartStereoProjectionFactor: triangulates point and keeps an estimate of it around. This factor operates with StereoCamera. This factor requires that values contains the involved StereoCameras. Calibration is assumed to be fixed, as this is also assumed in StereoCamera. If you'd like to store poses in values instead of cameras, use SmartStereoProjectionPoseFactor instead
|
inline |
Constructor
params | internal parameters of the smart factors |
|
inlineoverride |
Virtual destructor
|
inlinevirtualinherited |
Checks whether a factor should be used based on a set of values. This is primarily used to implement inequality constraints that require a variable active set. For all others, the default implementation returning true solves this problem.
In an inequality/bounding constraint, this active() returns true when the constraint is NOT fulfilled.
Reimplemented in gtsam::BoundingConstraint2< VALUE1, VALUE2 >, gtsam::AntiFactor, and gtsam::BoundingConstraint1< VALUE >.
|
inlineinherited |
Add a new measurement and pose/camera key.
measured | is the 2m dimensional projection of a single landmark |
key | is the index corresponding to the camera observing the landmark |
|
inlineinherited |
Add an entire SfM_track (collection of cameras observing a single point). The noise is assumed to be the same for all measurements.
|
inlineinherited |
Iterator at beginning of involved variable keys
|
inlineinherited |
Iterator at beginning of involved variable keys
|
inlinevirtualinherited |
Creates a shared_ptr clone of the factor - needs to be specialized to allow for subclasses
By default, throws exception if subclass does not implement the function.
Reimplemented in gtsam::EssentialMatrixFactor4< CALIBRATION >, gtsam::NonlinearEquality2< T >, gtsam::CombinedImuFactor, gtsam::ImuFactor2, gtsam::EssentialMatrixFactor3, gtsam::NonlinearEquality1< VALUE >, gtsam::GeneralSFMFactor2< CALIBRATION >, gtsam::ImuFactor, gtsam::LinearizedHessianFactor, gtsam::PendulumFactorPk1, gtsam::FunctorizedFactor2< R, T1, T2 >, gtsam::MagFactor3, gtsam::Pose3AttitudeFactor, gtsam::SmartRangeFactor, gtsam::NonlinearEquality< VALUE >, gtsam::AHRSFactor, gtsam::EssentialMatrixFactor2, gtsam::ExpressionFactor< T >, gtsam::ExpressionFactor< BearingRange< A1, A2 > >, gtsam::ExpressionFactor< double >, gtsam::MagFactor2, gtsam::PendulumFactorPk, gtsam::GPSFactor2, gtsam::ProjectionFactorRollingShutter, gtsam::RangeFactorWithTransform< A1, A2, T >, gtsam::LinearContainerFactor, gtsam::LinearizedJacobianFactor, gtsam::Rot3AttitudeFactor, gtsam::MagFactor1, gtsam::GenericProjectionFactor< POSE, LANDMARK, CALIBRATION >, gtsam::DiscreteEulerPoincareHelicopter, gtsam::MultiProjectionFactor< POSE, LANDMARK, CALIBRATION >, gtsam::GeneralSFMFactor< CAMERA, LANDMARK >, gtsam::ProjectionFactorPPP< POSE, LANDMARK, CALIBRATION >, gtsam::TransformBtwRobotsUnaryFactorEM< VALUE >, gtsam::GenericStereoFactor< POSE, LANDMARK >, gtsam::TriangulationFactor< CAMERA >, gtsam::RotateDirectionsFactor, gtsam::PendulumFactor2, gtsam::ReferenceFrameFactor< POINT, TRANSFORM >, gtsam::FunctorizedFactor< R, T >, gtsam::PartialPriorFactor< VALUE >, gtsam::FunctorizedFactor< Vector, ParameterMatrix< M > >, gtsam::FunctorizedFactor< double, BASIS::Parameters >, gtsam::FunctorizedFactor< double, Vector >, gtsam::FunctorizedFactor< T, ParameterMatrix< traits< T >::dimension > >, gtsam::FunctorizedFactor< double, ParameterMatrix< P > >, gtsam::PartialPriorFactor< PoseRTV >, gtsam::MagPoseFactor< POSE >, gtsam::ProjectionFactorPPPC< POSE, LANDMARK, CALIBRATION >, gtsam::TransformBtwRobotsUnaryFactor< VALUE >, gtsam::BetweenFactor< VALUE >, gtsam::EssentialMatrixFactor, gtsam::VelocityConstraint, gtsam::PriorFactor< VALUE >, gtsam::GPSFactor, gtsam::PoseToPointFactor< POSE, POINT >, gtsam::BarometricFactor, gtsam::EssentialMatrixConstraint, gtsam::DummyFactor, gtsam::PoseBetweenFactor< POSE >, gtsam::BearingRangeFactor< A1, A2, B, R >, gtsam::PosePriorFactor< POSE >, gtsam::MagFactor, gtsam::PoseTranslationPrior< POSE >, gtsam::PoseRotationPrior< POSE >, gtsam::FullIMUFactor< POSE >, gtsam::AntiFactor, gtsam::RangeFactor< A1, A2, T >, gtsam::IMUFactor< POSE >, gtsam::RelativeElevationFactor, gtsam::PendulumFactor1, gtsam::Reconstruction, gtsam::RotateFactor, and gtsam::VelocityConstraint3.
|
inlineinherited |
Compute F, E, and b (called below in both vanilla and SVD versions), where F is a vector of derivatives wrpt the cameras, and E the stacked derivatives with respect to the point. The value of cameras/point are passed as parameters.
|
inlineinherited |
SVD version that produces smaller Jacobian matrices by doing an SVD decomposition on E, and returning the left nulkl-space of E. See JacobianFactorSVD for more documentation.
|
inline |
Compute F, E only (called below in both vanilla and SVD versions) Assumes the point has been computed Note E can be 2m*3 or 2m*2, in case point is degenerate
|
inlineinherited |
An overload of correctForMissingMeasurements. This allows end users to provide optional arguments that are l-value references to the matrices and vectors that will be used to store the results instead of pointers.
|
inlineoverridevirtual |
This corrects the Jacobians and error vector for the case in which the right 2D measurement in the monocular camera is missing (nan).
Reimplemented from gtsam::SmartFactorBase< StereoCamera >.
|
inlineinherited |
Return Jacobians as JacobianFactorSVD. TODO(dellaert): lambda is currently ignored
|
inlineinherited |
Iterator at end of involved variable keys
|
inlineinherited |
Iterator at end of involved variable keys
|
overridevirtualinherited |
All factor types need to implement an error function. In factor graphs, this is the negative log-likelihood.
Reimplemented from gtsam::Factor.
|
inlinestaticprotectedinherited |
Construct factor from iterator keys. This is called internally from derived factor static factor methods, as a workaround for not being able to call the protected constructors above.
|
inlinestaticprotectedinherited |
Construct factor from container of keys. This is called internally from derived factor static factor methods, as a workaround for not being able to call the protected constructors above.
|
inline |
return the degenerate state
|
inline |
return the farPoint state
|
inline |
return the outlier state
|
inline |
return the cheirality status flag
|
inlineinherited |
|
inline |
|
inline |
|
inline |
return the landmark
|
inline |
COMPUTE the landmark
|
inlineoverridevirtual |
s | optional string naming the factor |
keyFormatter | optional formatter useful for printing Symbols |
Reimplemented from gtsam::NonlinearFactor.
Reimplemented in gtsam::SmartStereoProjectionFactorPP, and gtsam::SmartStereoProjectionPoseFactor.
|
virtualinherited |
Creates a shared_ptr clone of the factor with different keys using a map from old->new keys
Reimplemented in gtsam::LinearContainerFactor.
|
virtualinherited |
Clones a factor and fully replaces its keys
new_keys | is the full replacement set of keys |
Reimplemented in gtsam::LinearContainerFactor.
|
inlinevirtualinherited |
Should the factor be evaluated in the same thread as the caller This is to enable factors that has shared states (like the Python GIL lock)
Reimplemented in gtsam::CustomFactor.
|
inlineinherited |
|
inlineinherited |
Calculate the error of the factor. This is the log-likelihood, e.g. \( 0.5(h(x)-z)^2/\sigma^2 \) in case of Gaussian. In this class, we take the raw prediction error \( h(x)-z \), ask the noise model to transform it to \( (h(x)-z)^2/\sigma^2 \), and then multiply by 0.5. Will be used in "error(Values)" function required by NonlinearFactor base class
|
inline |
Calculate the error of the factor. This is the log-likelihood, e.g. \( 0.5(h(x)-z)^2/\sigma^2 \) in case of Gaussian. In this class, we take the raw prediction error \( h(x)-z \), ask the noise model to transform it to \( (h(x)-z)^2/\sigma^2 \), and then multiply by 0.5.
|
inline |
Triangulate and compute derivative of error with respect to point
|
inline |
Triangulate and compute derivative of error with respect to point
|
inlineinherited |
Compute reprojection errors [h(x)-z] = [cameras.project(p)-z] and derivatives. This is the error before the noise model is applied. The templated version described above must finally get resolved to this function.
|
inlineinherited |
An overload of unwhitenedError. This allows end users to provide optional arguments that are l-value references to the matrices and vectors that will be used to store the results instead of pointers.
|
inlineinherited |
Add the contribution of the smart factor to a pre-allocated Hessian, using sparse linear algebra. More efficient than the creation of the Hessian without preallocation of the SymmetricBlockMatrix
|
inlineinherited |
Calculate vector of re-projection errors [h(x)-z] = [cameras.project(p) - z], with the noise model applied.
|
protectedinherited |
Measurements for each of the m views. We keep a copy of the measurements for I/O and computing the error. The order is kept the same as the keys that we use to create the factor.
|
protectedinherited |
As of Feb 22, 2015, the noise model is the same for all measurements and is isotropic. This allows for moving most calculations of Schur complement etc. to be easily moved to CameraSet, and also agrees pragmatically with what is normally done.