Main Page   Class Overview   Pixels   Images   Geometry   Quick Index  

HxGaussDerivative3d.h File Reference

More...

#include "HxImageRep.h"

Go to the source code of this file.

Functions

HxImageRep L_HXIMAGEREP HxGaussDerivative3d (HxImageRep img, double sigma, int orderDerivx, int orderDerivy, int orderDerivz, double accuracy=3.0)
 Convolution Gaussian. More...


Detailed Description


Function Documentation

HxImageRep L_HXIMAGEREP HxGaussDerivative3d HxImageRep    img,
double    sigma,
int    orderDerivx,
int    orderDerivy,
int    orderDerivz,
double    accuracy = 3.0
 

Convolution Gaussian.

Equivalent to : img.genConvSeparated( gauss, "mul", "addAssign", HxImageRep::ARITH_PREC)

where gauss is the 1d double-precision Gaussian kernel based on sigma, orderDeriv, accuracy, for each dimension.

Notice that the kernel is applied to every dimension of the image separately and that the result image has a double-precision pixel type.

00018 {
00019     HxImageRep gaussx, gaussy, gaussz;
00020 
00021     gaussx = HxMakeGaussian1d(sigma, orderDerivx, accuracy,
00022         img.dimensionSize(1));
00023     gaussy = HxMakeGaussian1d(sigma, orderDerivy, accuracy,
00024         img.dimensionSize(2));
00025     gaussz = HxMakeGaussian1d(sigma, orderDerivz, accuracy,
00026         img.dimensionSize(3));
00027 
00028     HxImageRep res = img.generalizedConvolutionK1d(1, gaussx,
00029         "mul", "addAssign", HxImageRep::ARITH_PREC);
00030     res = res.generalizedConvolutionK1d(2, gaussy,
00031         "mul", "addAssign", HxImageRep::ARITH_PREC);
00032     return res.generalizedConvolutionK1d(3, gaussz,
00033         "mul", "addAssign", HxImageRep::ARITH_PREC);
00034 }


Generated on Tue Jan 8 13:59:20 2002 for C++Reference by doxygen1.2.12 written by Dimitri van Heesch, © 1997-2001