OpenCV  4.1.0
Open Source Computer Vision
Public Member Functions | Static Public Member Functions | List of all members
cv::cuda::StereoBeliefPropagation Class Referenceabstract

Class computing stereo correspondence using the belief propagation algorithm. : More...

#include <opencv2/cudastereo.hpp>

Inheritance diagram for cv::cuda::StereoBeliefPropagation:
cv::StereoMatcher cv::Algorithm cv::cuda::StereoConstantSpaceBP

Public Member Functions

virtual void compute (InputArray left, InputArray right, OutputArray disparity, Stream &stream)=0
 
virtual void compute (InputArray data, OutputArray disparity, Stream &stream=Stream::Null())=0
 Enables the stereo correspondence operator that finds the disparity for the specified data cost.
 
virtual double getDataWeight () const =0
 data weight
 
virtual double getDiscSingleJump () const =0
 discontinuity single jump
 
virtual double getMaxDataTerm () const =0
 truncation of data cost
 
virtual double getMaxDiscTerm () const =0
 truncation of discontinuity cost
 
virtual int getMsgType () const =0
 type for messages (CV_16SC1 or CV_32FC1)
 
virtual int getNumIters () const =0
 number of BP iterations on each level
 
virtual int getNumLevels () const =0
 number of levels
 
virtual void setDataWeight (double data_weight)=0
 
virtual void setDiscSingleJump (double disc_single_jump)=0
 
virtual void setMaxDataTerm (double max_data_term)=0
 
virtual void setMaxDiscTerm (double max_disc_term)=0
 
virtual void setMsgType (int msg_type)=0
 
virtual void setNumIters (int iters)=0
 
virtual void setNumLevels (int levels)=0
 
- Public Member Functions inherited from cv::StereoMatcher
virtual void compute (InputArray left, InputArray right, OutputArray disparity)=0
 Computes disparity map for the specified stereo pair.
 
virtual int getBlockSize () const =0
 
virtual int getDisp12MaxDiff () const =0
 
virtual int getMinDisparity () const =0
 
virtual int getNumDisparities () const =0
 
virtual int getSpeckleRange () const =0
 
virtual int getSpeckleWindowSize () const =0
 
virtual void setBlockSize (int blockSize)=0
 
virtual void setDisp12MaxDiff (int disp12MaxDiff)=0
 
virtual void setMinDisparity (int minDisparity)=0
 
virtual void setNumDisparities (int numDisparities)=0
 
virtual void setSpeckleRange (int speckleRange)=0
 
virtual void setSpeckleWindowSize (int speckleWindowSize)=0
 
- Public Member Functions inherited from cv::Algorithm
 Algorithm ()
 
virtual ~Algorithm ()
 
virtual void clear ()
 Clears the algorithm state.
 
virtual bool empty () const
 Returns true if the Algorithm is empty (e.g. in the very beginning or after unsuccessful read.
 
virtual String getDefaultName () const
 
virtual void read (const FileNode &fn)
 Reads algorithm parameters from a file storage.
 
virtual void save (const String &filename) const
 
virtual void write (FileStorage &fs) const
 Stores algorithm parameters in a file storage.
 
void write (const Ptr< FileStorage > &fs, const String &name=String()) const
 simplified API for language bindingsThis is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.
 

Static Public Member Functions

static void estimateRecommendedParams (int width, int height, int &ndisp, int &iters, int &levels)
 Uses a heuristic method to compute the recommended parameters ( ndisp, iters and levels ) for the specified image size ( width and height ).
 

Additional Inherited Members

- Public Types inherited from cv::StereoMatcher
enum  {
  DISP_SHIFT = 4,
  DISP_SCALE = (1 << DISP_SHIFT)
}
 
- Protected Member Functions inherited from cv::Algorithm
void writeFormat (FileStorage &fs) const
 

Detailed Description

Class computing stereo correspondence using the belief propagation algorithm. :

The class implements algorithm described inFelzenszwalb2006 . It can compute own data cost (using a truncated linear model) or use a user-provided data cost.

Note
StereoBeliefPropagation requires a lot of memory for message storage:

\[width \_ step \cdot height \cdot ndisp \cdot 4 \cdot (1 + 0.25)\]

and for data cost storage:

\[width\_step \cdot height \cdot ndisp \cdot (1 + 0.25 + 0.0625 + \dotsm + \frac{1}{4^{levels}})\]

width_step is the number of bytes in a line including padding.

StereoBeliefPropagation uses a truncated linear model for the data cost and discontinuity terms:

\[DataCost = data \_ weight \cdot \min ( \lvert Img_Left(x,y)-Img_Right(x-d,y) \rvert , max \_ data \_ term)\]

\[DiscTerm = \min (disc \_ single \_ jump \cdot \lvert f_1-f_2 \rvert , max \_ disc \_ term)\]

For more details, seeFelzenszwalb2006 .

By default, StereoBeliefPropagation uses floating-point arithmetics and the CV_32FC1 type for messages. But it can also use fixed-point arithmetics and the CV_16SC1 message type for better performance. To avoid an overflow in this case, the parameters must satisfy the following requirement:

\[10 \cdot 2^{levels-1} \cdot max \_ data \_ term < SHRT \_ MAX\]

See Also
StereoMatcher

Member Function Documentation

virtual void cv::cuda::StereoBeliefPropagation::compute ( InputArray  left,
InputArray  right,
OutputArray  disparity,
Stream stream 
)
pure virtual

This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

virtual void cv::cuda::StereoBeliefPropagation::compute ( InputArray  data,
OutputArray  disparity,
Stream stream = Stream::Null() 
)
pure virtual

Enables the stereo correspondence operator that finds the disparity for the specified data cost.

Parameters
dataUser-specified data cost, a matrix of msg_type type and Size(<image columns>*ndisp, <image rows>) size.
disparityOutput disparity map. If disparity is empty, the output type is CV_16SC1 . Otherwise, the type is retained. In 16-bit signed format, the disparity values do not have fractional bits.
streamStream for the asynchronous version.
static void cv::cuda::StereoBeliefPropagation::estimateRecommendedParams ( int  width,
int  height,
int &  ndisp,
int &  iters,
int &  levels 
)
static

Uses a heuristic method to compute the recommended parameters ( ndisp, iters and levels ) for the specified image size ( width and height ).

virtual double cv::cuda::StereoBeliefPropagation::getDataWeight ( ) const
pure virtual

data weight

virtual double cv::cuda::StereoBeliefPropagation::getDiscSingleJump ( ) const
pure virtual

discontinuity single jump

virtual double cv::cuda::StereoBeliefPropagation::getMaxDataTerm ( ) const
pure virtual

truncation of data cost

virtual double cv::cuda::StereoBeliefPropagation::getMaxDiscTerm ( ) const
pure virtual

truncation of discontinuity cost

virtual int cv::cuda::StereoBeliefPropagation::getMsgType ( ) const
pure virtual

type for messages (CV_16SC1 or CV_32FC1)

virtual int cv::cuda::StereoBeliefPropagation::getNumIters ( ) const
pure virtual

number of BP iterations on each level

virtual int cv::cuda::StereoBeliefPropagation::getNumLevels ( ) const
pure virtual

number of levels

virtual void cv::cuda::StereoBeliefPropagation::setDataWeight ( double  data_weight)
pure virtual
virtual void cv::cuda::StereoBeliefPropagation::setDiscSingleJump ( double  disc_single_jump)
pure virtual
virtual void cv::cuda::StereoBeliefPropagation::setMaxDataTerm ( double  max_data_term)
pure virtual
virtual void cv::cuda::StereoBeliefPropagation::setMaxDiscTerm ( double  max_disc_term)
pure virtual
virtual void cv::cuda::StereoBeliefPropagation::setMsgType ( int  msg_type)
pure virtual
virtual void cv::cuda::StereoBeliefPropagation::setNumIters ( int  iters)
pure virtual
virtual void cv::cuda::StereoBeliefPropagation::setNumLevels ( int  levels)
pure virtual

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