Package boofcv.alg.geo
Class NormalizationPoint2D
java.lang.Object
boofcv.alg.geo.NormalizationPoint2D
Describes how to normalize a set of points such that they have zero mean and variance. This is equivalent
to applying the matrix below. Normalization is often needed as a preprocessing step for solving linear equations.
Greatly reduces bias and numerical errors.
N = [ 1/σ_x 0 -μ_x/σ_x ] [ 0 1/σ_y 0 -μ_y/σ_y ] [ 0 0 1 ]
Y. Ma, S. Soatto, J. Kosecka, and S. S. Sastry, "An Invitation to 3-D Vision" Springer-Verlad, 2004
-
Field Summary
-
Constructor Summary
-
Method Summary
Modifier and TypeMethodDescriptionvoid
apply
(ConicGeneral_F64 p, ConicGeneral_F64 output) C* = H'*C*Hvoid
apply
(Point2D_F64 p, Point2D_F64 output) void
apply
(Point3D_F64 p, Point3D_F64 output) void
apply
(DMatrix3x3 C, DMatrix3x3 output) Apply transform to conic in 3x3 matrix format.void
apply
(DMatrixRMaj H, DMatrixRMaj output) Applies normalization to a H=3xN matrix out = Norm*Hboolean
isEquals
(NormalizationPoint2D a, double tol) matrix
(@Nullable DMatrixRMaj M) matrix3
(@Nullable DMatrix3x3 M) matrixInv
(@Nullable DMatrixRMaj M) void
remove
(ConicGeneral_F64 p, ConicGeneral_F64 output) void
remove
(Point2D_F64 p, Point2D_F64 output) void
remove
(Point3D_F64 p, Point3D_F64 output) void
remove
(DMatrix3x3 C, DMatrix3x3 output) void
remove
(DMatrixRMaj H, DMatrixRMaj output) Applies normalization to a H=3xN matrix out = Norm*Hvoid
set
(double meanX, double stdX, double meanY, double stdY)
-
Field Details
-
meanX
public double meanX -
stdX
public double stdX -
meanY
public double meanY -
stdY
public double stdY
-
-
Constructor Details
-
NormalizationPoint2D
public NormalizationPoint2D() -
NormalizationPoint2D
public NormalizationPoint2D(double meanX, double stdX, double meanY, double stdY)
-
-
Method Details
-
set
public void set(double meanX, double stdX, double meanY, double stdY) -
apply
Applies normalization to a H=3xN matrix out = Norm*H- Parameters:
H
- 3xN matrix. Can be same as input matrix
-
remove
Applies normalization to a H=3xN matrix out = Norm*H- Parameters:
H
- 3xN matrix. Can be same as input matrix
-
apply
-
apply
-
apply
C* = H'*C*H -
apply
Apply transform to conic in 3x3 matrix format. -
remove
-
remove
-
remove
-
remove
-
matrix
-
matrixInv
-
matrix3
-
matrixInv3
-
isEquals
-