cv::sfm::SFMLibmvEuclideanReconstruction Class Reference

SFMLibmvEuclideanReconstruction class provides an interface with the Libmv Structure From Motion pipeline. More...

#include <opencv2/sfm/simple_pipeline.hpp>

Inheritance diagram for cv::sfm::SFMLibmvEuclideanReconstruction:

Public Member Functions
virtual void	getCameras (OutputArray Rs, OutputArray Ts) CV_OVERRIDE=0
	Returns the estimated camera extrinsic parameters. More...
virtual double	getError () const CV_OVERRIDE=0
	Returns the computed reprojection error. More...
virtual cv::Mat	getIntrinsics () const CV_OVERRIDE=0
	Returns the refined camera calibration matrix. More...
virtual void	getPoints (OutputArray points3d) CV_OVERRIDE=0
	Returns the estimated 3d points. More...
virtual void	run (InputArrayOfArrays points2d) CV_OVERRIDE=0
	Calls the pipeline in order to perform Eclidean reconstruction. More...
virtual void	run (InputArrayOfArrays points2d, InputOutputArray K, OutputArray Rs, OutputArray Ts, OutputArray points3d) CV_OVERRIDE=0
	Calls the pipeline in order to perform Eclidean reconstruction. More...
virtual void	run (const std::vector< String > &images) CV_OVERRIDE=0
	Calls the pipeline in order to perform Eclidean reconstruction. More...
virtual void	run (const std::vector< String > &images, InputOutputArray K, OutputArray Rs, OutputArray Ts, OutputArray points3d) CV_OVERRIDE=0
	Calls the pipeline in order to perform Eclidean reconstruction. More...
virtual void	setCameraIntrinsicOptions (const libmv_CameraIntrinsicsOptions &libmv_camera_intrinsics_options) CV_OVERRIDE=0
	Setter method for camera intrinsic options. More...
virtual void	setReconstructionOptions (const libmv_ReconstructionOptions &libmv_reconstruction_options) CV_OVERRIDE=0
	Setter method for reconstruction options. More...
Public Member Functions inherited from cv::sfm::BaseSFM
virtual	~BaseSFM ()

Static Public Member Functions
static Ptr< SFMLibmvEuclideanReconstruction >	create (const libmv_CameraIntrinsicsOptions &camera_instrinsic_options=libmv_CameraIntrinsicsOptions(), const libmv_ReconstructionOptions &reconstruction_options=libmv_ReconstructionOptions())
	Creates an instance of the SFMLibmvEuclideanReconstruction class. Initializes Libmv. More...

Detailed Description

SFMLibmvEuclideanReconstruction class provides an interface with the Libmv Structure From Motion pipeline.

Member Function Documentation

create()

static Ptr<SFMLibmvEuclideanReconstruction> cv::sfm::SFMLibmvEuclideanReconstruction::create ( const libmv_CameraIntrinsicsOptions &  camera_instrinsic_options = libmv_CameraIntrinsicsOptions(), const libmv_ReconstructionOptions &  reconstruction_options = libmv_ReconstructionOptions()  )

static

Creates an instance of the SFMLibmvEuclideanReconstruction class. Initializes Libmv.

getCameras()

virtual void cv::sfm::SFMLibmvEuclideanReconstruction::getCameras ( OutputArray  Rs, OutputArray  Ts  )

pure virtual

Returns the estimated camera extrinsic parameters.

Parameters

Rs	Output vector of 3x3 rotations of the camera.
Ts	Output vector of 3x1 translations of the camera.

Implements cv::sfm::BaseSFM.

getError()

virtual double cv::sfm::SFMLibmvEuclideanReconstruction::getError ( ) const

pure virtual

Returns the computed reprojection error.

Implements cv::sfm::BaseSFM.

getIntrinsics()

virtual cv::Mat cv::sfm::SFMLibmvEuclideanReconstruction::getIntrinsics ( ) const

pure virtual

Returns the refined camera calibration matrix.

Implements cv::sfm::BaseSFM.

getPoints()

virtual void cv::sfm::SFMLibmvEuclideanReconstruction::getPoints ( OutputArray  points3d )

pure virtual

Returns the estimated 3d points.

Parameters

points3d

Output array with estimated 3d points.

Implements cv::sfm::BaseSFM.

run() [1/4]

virtual void cv::sfm::SFMLibmvEuclideanReconstruction::run ( InputArrayOfArrays  points2d )

pure virtual

Calls the pipeline in order to perform Eclidean reconstruction.

Parameters

points2d

Input vector of vectors of 2d points (the inner vector is per image).

Note

• Tracks must be as precise as possible. It does not handle outliers and is very sensible to them.

Implements cv::sfm::BaseSFM.

run() [2/4]

virtual void cv::sfm::SFMLibmvEuclideanReconstruction::run ( InputArrayOfArrays  points2d, InputOutputArray  K, OutputArray  Rs, OutputArray  Ts, OutputArray  points3d  )

pure virtual

Calls the pipeline in order to perform Eclidean reconstruction.

Parameters

points2d	Input vector of vectors of 2d points (the inner vector is per image).
K	Input/Output camera matrix \(K = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{1}\). Input parameters used as initial guess.
Rs	Output vector of 3x3 rotations of the camera.
Ts	Output vector of 3x1 translations of the camera.
points3d	Output array with estimated 3d points.

Note

• Tracks must be as precise as possible. It does not handle outliers and is very sensible to them.

Implements cv::sfm::BaseSFM.

run() [3/4]

virtual void cv::sfm::SFMLibmvEuclideanReconstruction::run ( const std::vector< String > &  images )

pure virtual

Calls the pipeline in order to perform Eclidean reconstruction.

Parameters

images

a vector of string with the images paths.

Note

• The images must be ordered as they were an image sequence. Additionally, each frame should be as close as posible to the previous and posterior.
• For now DAISY features are used in order to compute the 2d points tracks and it only works for 3-4 images.

Implements cv::sfm::BaseSFM.

run() [4/4]

virtual void cv::sfm::SFMLibmvEuclideanReconstruction::run ( const std::vector< String > &  images, InputOutputArray  K, OutputArray  Rs, OutputArray  Ts, OutputArray  points3d  )

pure virtual

Calls the pipeline in order to perform Eclidean reconstruction.

Parameters

images	a vector of string with the images paths.
K	Input/Output camera matrix \(K = \vecthreethree{f_x}{0}{c_x}{0}{f_y}{c_y}{0}{0}{1}\). Input parameters used as initial guess.
Rs	Output vector of 3x3 rotations of the camera.
Ts	Output vector of 3x1 translations of the camera.
points3d	Output array with estimated 3d points.

Note

• The images must be ordered as they were an image sequence. Additionally, each frame should be as close as posible to the previous and posterior.
• For now DAISY features are used in order to compute the 2d points tracks and it only works for 3-4 images.

Implements cv::sfm::BaseSFM.

setCameraIntrinsicOptions()

virtual void cv::sfm::SFMLibmvEuclideanReconstruction::setCameraIntrinsicOptions ( const libmv_CameraIntrinsicsOptions &  libmv_camera_intrinsics_options )

pure virtual

Setter method for camera intrinsic options.

Parameters

libmv_camera_intrinsics_options

struct with camera intrinsic options such as camera model and the internal camera parameters.

Implements cv::sfm::BaseSFM.

setReconstructionOptions()

virtual void cv::sfm::SFMLibmvEuclideanReconstruction::setReconstructionOptions ( const libmv_ReconstructionOptions &  libmv_reconstruction_options )

pure virtual

Setter method for reconstruction options.

Parameters

libmv_reconstruction_options

struct with reconstruction options such as initial keyframes, automatic keyframe selection, parameters to refine and the verbosity level.

Implements cv::sfm::BaseSFM.

---

The documentation for this class was generated from the following file:

• opencv2/sfm/simple_pipeline.hpp