Shape Transformers and Interfaces

A virtual interface that ease the use of transforming algorithms in some pipelines, such as the Shape Context Matching Algorithm. Thus, all objects that implement shape transformation techniques inherit the ShapeTransformer interface.

ShapeTransformer

class ShapeTransformer : public Algorithm

Abstract base class for shape transformation algorithms.

class CV_EXPORTS_W ShapeTransformer : public Algorithm
{
public:
    CV_WRAP virtual void estimateTransformation(InputArray transformingShape, InputArray targetShape,
                                                 std::vector<DMatch>& matches) = 0;

    CV_WRAP virtual float applyTransformation(InputArray input, OutputArray output=noArray()) = 0;

    CV_WRAP virtual void warpImage(InputArray transformingImage, OutputArray output,
                                   int flags=INTER_LINEAR, int borderMode=BORDER_CONSTANT,
                                   const Scalar& borderValue=Scalar()) const = 0;
};

ShapeTransformer::estimateTransformation

Estimate the transformation parameters of the current transformer algorithm, based on point matches.

C++: void estimateTransformation(InputArray transformingShape, InputArray targetShape, std::vector<DMatch>& matches)
Parameters:
  • transformingShape – Contour defining first shape.
  • targetShape – Contour defining second shape (Target).
  • matches – Standard vector of Matches between points.

ShapeTransformer::applyTransformation

Apply a transformation, given a pre-estimated transformation parameters.

C++: float applyTransformation(InputArray input, OutputArray output=noArray() )
Parameters:
  • input – Contour (set of points) to apply the transformation.
  • output – Output contour.

ShapeTransformer::warpImage

Apply a transformation, given a pre-estimated transformation parameters, to an Image.

C++: void warpImage(InputArray transformingImage, OutputArray output, int flags=INTER_LINEAR, int borderMode=BORDER_CONSTANT, const Scalar& borderValue=Scalar() )
Parameters:
  • transformingImage – Input image.
  • output – Output image.
  • flags – Image interpolation method.
  • borderMode – border style.
  • borderValue – border value.

ThinPlateSplineShapeTransformer

class ThinPlateSplineShapeTransformer : public Algorithm

Definition of the transformation ocupied in the paper “Principal Warps: Thin-Plate Splines and Decomposition of Deformations”, by F.L. Bookstein (PAMI 1989).

class CV_EXPORTS_W ThinPlateSplineShapeTransformer : public ShapeTransformer
{
public:
    CV_WRAP virtual void setRegularizationParameter(double beta) = 0;
    CV_WRAP virtual double getRegularizationParameter() const = 0;
};

/* Complete constructor */
CV_EXPORTS_W Ptr<ThinPlateSplineShapeTransformer>
    createThinPlateSplineShapeTransformer(double regularizationParameter=0);

ThinPlateSplineShapeTransformer::setRegularizationParameter

Set the regularization parameter for relaxing the exact interpolation requirements of the TPS algorithm.

C++: void setRegularizationParameter(double beta)
Parameters:
  • beta – value of the regularization parameter.

AffineTransformer

class AffineTransformer : public Algorithm

Wrapper class for the OpenCV Affine Transformation algorithm.

class CV_EXPORTS_W AffineTransformer : public ShapeTransformer
{
public:
    CV_WRAP virtual void setFullAffine(bool fullAffine) = 0;
    CV_WRAP virtual bool getFullAffine() const = 0;
};

/* Complete constructor */
CV_EXPORTS_W Ptr<AffineTransformer> createAffineTransformer(bool fullAffine);