|
<html><head><title>dlib C++ Library - hessian_pyramid_abstract.h</title></head><body bgcolor='white'><pre> |
|
<font color='#009900'>// Copyright (C) 2009 Davis E. King ([email protected]) |
|
</font><font color='#009900'>// License: Boost Software License See LICENSE.txt for the full license. |
|
</font><font color='#0000FF'>#undef</font> DLIB_HESSIAN_PYRAMId_ABSTRACT_Hh_ |
|
<font color='#0000FF'>#ifdef</font> DLIB_HESSIAN_PYRAMId_ABSTRACT_Hh_ |
|
|
|
<font color='#0000FF'>#include</font> "<a style='text-decoration:none' href='../image_transforms/integral_image_abstract.h.html'>../image_transforms/integral_image_abstract.h</a>" |
|
<font color='#0000FF'>#include</font> "<a style='text-decoration:none' href='../noncopyable.h.html'>../noncopyable.h</a>" |
|
<font color='#0000FF'>#include</font> <font color='#5555FF'><</font>vector<font color='#5555FF'>></font> |
|
|
|
<font color='#0000FF'>namespace</font> dlib |
|
<b>{</b> |
|
|
|
<font color='#0000FF'>class</font> <b><a name='hessian_pyramid'></a>hessian_pyramid</b> : noncopyable |
|
<b>{</b> |
|
<font color='#009900'>/*! |
|
INITIAL VALUE |
|
- octaves() == 0 |
|
- intervals() == 0 |
|
|
|
WHAT THIS OBJECT REPRESENTS |
|
This object represents an image pyramid where each level in the |
|
pyramid holds determinants of Hessian matrices for the original |
|
input image. This object can be used to find stable interest |
|
points in an image. For further details consult the following |
|
papers. |
|
|
|
This object is an implementation of the fast Hessian pyramid |
|
as described in the paper: |
|
SURF: Speeded Up Robust Features |
|
By Herbert Bay, Tinne Tuytelaars, and Luc Van Gool |
|
|
|
This implementation was also influenced by the very well documented |
|
OpenSURF library and its corresponding description of how the fast |
|
Hessian algorithm functions: |
|
Notes on the OpenSURF Library |
|
Christopher Evans |
|
!*/</font> |
|
<font color='#0000FF'>public</font>: |
|
|
|
<font color='#0000FF'>template</font> <font color='#5555FF'><</font><font color='#0000FF'>typename</font> integral_image_type<font color='#5555FF'>></font> |
|
<font color='#0000FF'><u>void</u></font> <b><a name='build_pyramid'></a>build_pyramid</b> <font face='Lucida Console'>(</font> |
|
<font color='#0000FF'>const</font> integral_image_type<font color='#5555FF'>&</font> img, |
|
<font color='#0000FF'><u>long</u></font> num_octaves, |
|
<font color='#0000FF'><u>long</u></font> num_intervals, |
|
<font color='#0000FF'><u>long</u></font> initial_step_size |
|
<font face='Lucida Console'>)</font>; |
|
<font color='#009900'>/*! |
|
requires |
|
- num_octaves > 0 |
|
- num_intervals > 0 |
|
- initial_step_size > 0 |
|
- integral_image_type == an object such as dlib::integral_image or another |
|
type that implements the interface defined in image_transforms/integral_image_abstract.h |
|
ensures |
|
- #get_step_size(0) == initial_step_size |
|
- #octaves() == num_octaves |
|
- #intervals() == num_intervals |
|
- creates a Hessian pyramid from the given input image. |
|
!*/</font> |
|
|
|
<font color='#0000FF'><u>long</u></font> <b><a name='octaves'></a>octaves</b> <font face='Lucida Console'>(</font> |
|
<font face='Lucida Console'>)</font> <font color='#0000FF'>const</font>; |
|
<font color='#009900'>/*! |
|
ensures |
|
- returns the number of octaves in this pyramid |
|
!*/</font> |
|
|
|
<font color='#0000FF'><u>long</u></font> <b><a name='intervals'></a>intervals</b> <font face='Lucida Console'>(</font> |
|
<font face='Lucida Console'>)</font> <font color='#0000FF'>const</font>; |
|
<font color='#009900'>/*! |
|
ensures |
|
- returns the number of intervals in this pyramid |
|
!*/</font> |
|
|
|
<font color='#0000FF'><u>long</u></font> <b><a name='get_border_size'></a>get_border_size</b> <font face='Lucida Console'>(</font> |
|
<font color='#0000FF'><u>long</u></font> interval |
|
<font face='Lucida Console'>)</font> <font color='#0000FF'>const</font>; |
|
<font color='#009900'>/*! |
|
requires |
|
- 0 <= interval < intervals() |
|
ensures |
|
- Each interval of the pyramid has a certain sized border region where we |
|
can't compute the Hessian values since they are too close to the edge |
|
of the input image. This function returns the size of that border. |
|
!*/</font> |
|
|
|
<font color='#0000FF'><u>long</u></font> <b><a name='get_step_size'></a>get_step_size</b> <font face='Lucida Console'>(</font> |
|
<font color='#0000FF'><u>long</u></font> octave |
|
<font face='Lucida Console'>)</font> <font color='#0000FF'>const</font>; |
|
<font color='#009900'>/*! |
|
requires |
|
- 0 <= octave < octaves() |
|
ensures |
|
- Each octave has a step size value. This value determines how many |
|
input image pixels separate each pixel in the given pyramid octave. |
|
As the octave gets larger (i.e. as it goes to the top of the pyramid) the |
|
step size gets bigger and thus the pyramid narrows. |
|
!*/</font> |
|
|
|
<font color='#0000FF'><u>long</u></font> <b><a name='nr'></a>nr</b> <font face='Lucida Console'>(</font> |
|
<font color='#0000FF'><u>long</u></font> octave |
|
<font face='Lucida Console'>)</font> <font color='#0000FF'>const</font>; |
|
<font color='#009900'>/*! |
|
requires |
|
- 0 <= octave < octaves() |
|
ensures |
|
- returns the number of rows there are per layer in the given |
|
octave of pyramid |
|
!*/</font> |
|
|
|
<font color='#0000FF'><u>long</u></font> <b><a name='nc'></a>nc</b> <font face='Lucida Console'>(</font> |
|
<font color='#0000FF'><u>long</u></font> octave |
|
<font face='Lucida Console'>)</font> <font color='#0000FF'>const</font>; |
|
<font color='#009900'>/*! |
|
requires |
|
- 0 <= octave < octaves() |
|
ensures |
|
- returns the number of columns there are per layer in the given |
|
octave of pyramid |
|
!*/</font> |
|
|
|
<font color='#0000FF'><u>double</u></font> <b><a name='get_value'></a>get_value</b> <font face='Lucida Console'>(</font> |
|
<font color='#0000FF'><u>long</u></font> octave, |
|
<font color='#0000FF'><u>long</u></font> interval, |
|
<font color='#0000FF'><u>long</u></font> r, |
|
<font color='#0000FF'><u>long</u></font> c |
|
<font face='Lucida Console'>)</font> <font color='#0000FF'>const</font>; |
|
<font color='#009900'>/*! |
|
requires |
|
- 0 <= octave < octaves() |
|
- 0 <= interval < intervals() |
|
- Let BS == get_border_size(interval): then |
|
- BS <= r < nr(octave)-BS |
|
- BS <= c < nc(octave)-BS |
|
ensures |
|
- returns the determinant of the Hessian from the given octave and interval |
|
of the pyramid. The specific point sampled at this pyramid level is |
|
the one that corresponds to the input image point (point(r,c)*get_step_size(octave)). |
|
!*/</font> |
|
|
|
<font color='#0000FF'><u>double</u></font> <b><a name='get_laplacian'></a>get_laplacian</b> <font face='Lucida Console'>(</font> |
|
<font color='#0000FF'><u>long</u></font> octave, |
|
<font color='#0000FF'><u>long</u></font> interval, |
|
<font color='#0000FF'><u>long</u></font> r, |
|
<font color='#0000FF'><u>long</u></font> c |
|
<font face='Lucida Console'>)</font> <font color='#0000FF'>const</font>; |
|
<font color='#009900'>/*! |
|
requires |
|
- 0 <= octave < octaves() |
|
- 0 <= interval < intervals() |
|
- Let BS == get_border_size(interval): then |
|
- BS <= r < nr(octave)-BS |
|
- BS <= c < nc(octave)-BS |
|
ensures |
|
- returns the sign of the laplacian for the given octave and interval |
|
of the pyramid. The specific point sampled at this pyramid level is |
|
the one that corresponds to the input image point (point(r,c)*get_step_size(octave)). |
|
- The laplacian is the trace of the Hessian at the given point. So this |
|
function returns either +1 or -1 depending on this number's sign. This |
|
value can be used to distinguish bright blobs on dark backgrounds from |
|
the reverse. |
|
!*/</font> |
|
|
|
<b>}</b>; |
|
|
|
<font color='#009900'>// ---------------------------------------------------------------------------------------- |
|
</font> |
|
<font color='#0000FF'>struct</font> <b><a name='interest_point'></a>interest_point</b> |
|
<b>{</b> |
|
<font color='#009900'>/*! |
|
WHAT THIS OBJECT REPRESENTS |
|
This object contains the interest points found using the |
|
hessian_pyramid object. Its fields have the following |
|
meanings: |
|
- center == the x/y location of the center of the interest point |
|
(in image space coordinates. y gives the row and x gives the |
|
column in the image) |
|
- scale == the scale at which the point was detected. This is a number |
|
>= 1. If it is 1 then it means the interest point was detected at |
|
the lowest scale in the image pyramid. Larger numbers indicate that |
|
the interest point is from high up in the image pyramid. For |
|
example, a scale of 4 would mean the interest point was located at a |
|
point in the pyramid where the image had been shrunk by a factor of 4. |
|
- score == the determinant of the Hessian for the interest point |
|
- laplacian == the sign of the laplacian for the interest point |
|
!*/</font> |
|
|
|
<b><a name='interest_point'></a>interest_point</b><font face='Lucida Console'>(</font><font face='Lucida Console'>)</font> : scale<font face='Lucida Console'>(</font><font color='#979000'>0</font><font face='Lucida Console'>)</font>, score<font face='Lucida Console'>(</font><font color='#979000'>0</font><font face='Lucida Console'>)</font>, laplacian<font face='Lucida Console'>(</font><font color='#979000'>0</font><font face='Lucida Console'>)</font> <b>{</b><b>}</b> |
|
|
|
dlib::vector<font color='#5555FF'><</font><font color='#0000FF'><u>double</u></font>,<font color='#979000'>2</font><font color='#5555FF'>></font> center; |
|
<font color='#0000FF'><u>double</u></font> scale; |
|
<font color='#0000FF'><u>double</u></font> score; |
|
<font color='#0000FF'><u>double</u></font> laplacian; |
|
|
|
<font color='#0000FF'><u>bool</u></font> <b><a name='operator'></a>operator</b> <font color='#5555FF'><</font> <font face='Lucida Console'>(</font><font color='#0000FF'>const</font> interest_point<font color='#5555FF'>&</font> p<font face='Lucida Console'>)</font> <font color='#0000FF'>const</font> <b>{</b> <font color='#0000FF'>return</font> score <font color='#5555FF'><</font> p.score; <b>}</b> |
|
<font color='#009900'>/*! |
|
This function is here so you can sort interest points according to |
|
their scores |
|
!*/</font> |
|
<b>}</b>; |
|
|
|
<font color='#009900'>// ---------------------------------------------------------------------------------------- |
|
</font> |
|
<font color='#0000FF'><u>void</u></font> <b><a name='serialize'></a>serialize</b> <font face='Lucida Console'>(</font> |
|
<font color='#0000FF'>const</font> interest_point<font color='#5555FF'>&</font> item, |
|
std::ostream<font color='#5555FF'>&</font> out |
|
<font face='Lucida Console'>)</font>; |
|
<font color='#009900'>/*! |
|
provides serialization support |
|
!*/</font> |
|
|
|
<font color='#0000FF'><u>void</u></font> <b><a name='deserialize'></a>deserialize</b> <font face='Lucida Console'>(</font> |
|
interest_point<font color='#5555FF'>&</font> item, |
|
std::istream<font color='#5555FF'>&</font> in |
|
<font face='Lucida Console'>)</font>; |
|
<font color='#009900'>/*! |
|
provides serialization support |
|
!*/</font> |
|
|
|
<font color='#009900'>// ---------------------------------------------------------------------------------------- |
|
</font> |
|
<font color='#0000FF'>template</font> <font color='#5555FF'><</font><font color='#0000FF'>typename</font> Alloc<font color='#5555FF'>></font> |
|
<font color='#0000FF'><u>void</u></font> <b><a name='get_interest_points'></a>get_interest_points</b> <font face='Lucida Console'>(</font> |
|
<font color='#0000FF'>const</font> hessian_pyramid<font color='#5555FF'>&</font> pyr, |
|
<font color='#0000FF'><u>double</u></font> threshold, |
|
std::vector<font color='#5555FF'><</font>interest_point,Alloc<font color='#5555FF'>></font><font color='#5555FF'>&</font> result_points |
|
<font face='Lucida Console'>)</font> |
|
<font color='#009900'>/*! |
|
requires |
|
- threshold >= 0 |
|
ensures |
|
- extracts interest points from the pyramid pyr and stores them into |
|
result_points (note that result_points is cleared before these new interest |
|
points are added to it). |
|
- Only interest points with determinant values in the pyramid larger than |
|
threshold are output. |
|
!*/</font> |
|
|
|
<font color='#009900'>// ---------------------------------------------------------------------------------------- |
|
</font> |
|
<b>}</b> |
|
|
|
<font color='#0000FF'>#endif</font> <font color='#009900'>// DLIB_HESSIAN_PYRAMId_ABSTRACT_Hh_ |
|
</font> |
|
|
|
</pre></body></html> |
