firevision::FilterLaplace Class Reference

Laplacian filter. More...

#include <filters/laplace.h>

Inheritance diagram for firevision::FilterLaplace:
Inheritance graph
[legend]

List of all members.

Public Member Functions

 FilterLaplace ()
 Constructor.
 FilterLaplace (float sigma, unsigned int size, float scale)
 Constructor.
 ~FilterLaplace ()
 Destructor.
virtual void apply ()
 Apply the filter.

Static Public Member Functions

static void calculate_kernel (int *kernel_buffer, float sigma, unsigned int size, float scale)
 Calculate a Laplacian of Gaussian kernel.

Detailed Description

Laplacian filter.

Laplacian of Gaussian filter.

Author:
Tim Niemueller

Definition at line 38 of file laplace.h.

Constructor & Destructor Documentation

firevision::FilterLaplace::FilterLaplace (  ) 

Constructor.

Definition at line 44 of file laplace.cpp.

firevision::FilterLaplace::FilterLaplace ( float  sigma,
unsigned int  size,
float  scale 
)

Constructor.

Parameters:
sigma sigma for Laplacian
size size of kernel
scale scale factor

Definition at line 56 of file laplace.cpp.

References calculate_kernel().

firevision::FilterLaplace::~FilterLaplace (  ) 

Destructor.

Definition at line 66 of file laplace.cpp.

Member Function Documentation

void firevision::FilterLaplace::apply (  )  [virtual]

Apply the filter.

Apply the filter to the given source and destination buffers with given width and height and orientation (ori may be ignored for some filters).

Implements firevision::Filter.

Definition at line 75 of file laplace.cpp.

References firevision::Filter::dst, firevision::Filter::dst_roi, firevision::ROI::height, firevision::ROI::line_step, firevision::ROI::pixel_step, firevision::Filter::src, firevision::Filter::src_roi, firevision::ROI::start, firevision::ROI::width, fawkes::point_t::x, and fawkes::point_t::y.

void firevision::FilterLaplace::calculate_kernel ( int *  kernel,
float  sigma,
unsigned int  size,
float  scale 
) [static]

Calculate a Laplacian of Gaussian kernel.

The kernel is calculated with the formula

\[ roundf( \frac{-1}{\pi * \sigma^4} * ( 1 - \frac{w^2 + h^2}{2 * \sigma^2} ) * e^{-\frac{w^2 + h^2}{2 * \sigma^2}} * \mathtt{scale} ) \]

Parameters:
kernel buffer contains kernel upon return
sigma sigma for formula
size kernel is of quadratic size $\mathtt{size} \times \mathtt{size}$
scale scale parameter in formula

Definition at line 148 of file laplace.cpp.

Referenced by FilterLaplace().

The documentation for this class was generated from the following files:
Generated on 1 Mar 2011 for Fawkes API by  doxygen 1.6.1