class PatternDetect

This class is designed to scan an image at several scales and locations

Inheritance:


Public Fields

[more]PatternCollector* patterncollector
the pattern collector
[more]int n_patterns
number of output patterns (detected and merged)
[more]Pattern** patterns
output patterns (detected and merged)
[more]int n_subwindows
total number of hypothesis
[more]int n_subwindows_processed
number of hypothesis tested
[more]int n_pruning_reject
number of hypothesis rejected by prunning
[more]int n_mean_reject
number of hypothesis rejected by mean
[more]int n_stdv_reject
number of hypothesis rejected by stdv

Public Methods

[more] PatternDetect( int width_image_, int height_image_, int patternWmin_, int patternHmin_, int Wmin_, int Wmax_, real stepx_, real stepy_, real scale_factor_, real minMean_, real maxMean_, real minStdv_, real maxStdv_, real* pruning_ii_, real pruning_min_threshold_, real pruning_max_threshold_, int max_pattern=10000, bool verbose=false)
detect patterns in an image using an ipSubwindow
[more]virtual void init(void)
[more]virtual void process(Sequence* seq_in)
[more]virtual void stat(void)
[more]int getWmin()
[more]int getWmax()
[more]virtual void mergepatterns(void) = 0
[more]virtual ~PatternDetect()


Documentation

This class is designed to scan an image at several scales and locations

    
	       width_image
	+----------------------+
	|   Wmin               |
	|  +-+                 |
	|  | |                 |
	|  | |         Wmax    |
	|  +-+      +--+       | height_image
	|           |  |       |
	|           |  |       |
	|           |  |       |
	|           +--+       |
	+----------------------+

    

oPatternCollector* patterncollector
the pattern collector

oint n_patterns
number of output patterns (detected and merged)

oPattern** patterns
output patterns (detected and merged)

oint n_subwindows
total number of hypothesis

oint n_subwindows_processed
number of hypothesis tested

oint n_pruning_reject
number of hypothesis rejected by prunning

oint n_mean_reject
number of hypothesis rejected by mean

oint n_stdv_reject
number of hypothesis rejected by stdv

o PatternDetect( int width_image_, int height_image_, int patternWmin_, int patternHmin_, int Wmin_, int Wmax_, real stepx_, real stepy_, real scale_factor_, real minMean_, real maxMean_, real minStdv_, real maxStdv_, real* pruning_ii_, real pruning_min_threshold_, real pruning_max_threshold_, int max_pattern=10000, bool verbose=false)
detect patterns in an image using an ipSubwindow

Parameters:
width_image_ - is the width of the image
height_image_ - is the height of the image
patternWmin_ - is the minimum width of the pattern
patternHmin_ - is the minimum height of the pattern
Wmin_ - is the minimum width of the subwindow to scan
Wmax_ - is the maximum width of the subwindow to scan
stepx_ - is the x step for scanning
stepy_ - is the y step for scanning
scale_factor_ - is the scale factor
minMean_ - is the minimum mean value of the pixels in a subwindow
maxMean_ - is the maximum mean value of the pixels in a subwindow
minStdv_ - is the minimum stdv value of the pixels in a subwindow
maxStdv_ - is the maximum stdv value of the pixels in a subwindow
pruning_ii_ - is the integral image of a binary pruning mask
pruning_min_threshold_ - is the min number of 1 in pruning_ii_
pruning_max_threshold_ - is the max number of 1 in pruning_ii_
max_pattern - is the maximum number of patterns that can be stored

ovirtual void init(void)

ovirtual void process(Sequence* seq_in)

ovirtual void stat(void)

oint getWmin()

oint getWmax()

ovirtual void mergepatterns(void) = 0

ovirtual ~PatternDetect()


Direct child classes:
MlpPatternDetect
MlpCascadePatternDetect
HaarPatternDetect
Author:
Sebastien Marcel (marcel@idiap.ch) Yann Rodriguez (rodrig@idiap.ch) Fri 15 Jul 2005 11:35:24 AM CEST
Version:
2.0
Since:
2.0

Alphabetic index HTML hierarchy of classes or Java



This page was generated with the help of DOC++.