OpenCV  2.4.13.2
Open Source Computer Vision
cv::linemod::QuantizedPyramid Class Referenceabstract

Represents a modality operating over an image pyramid. More...

#include <objdetect.hpp>

Classes

struct  Candidate
 Candidate feature with a score. More...
 

Public Member Functions

virtual ~QuantizedPyramid ()
 
virtual void quantize (Mat &dst) const =0
 Compute quantized image at current pyramid level for online detection. More...
 
virtual bool extractTemplate (Template &templ) const =0
 Extract most discriminant features at current pyramid level to form a new template. More...
 
virtual void pyrDown ()=0
 Go to the next pyramid level. More...
 

Static Protected Member Functions

static void selectScatteredFeatures (const std::vector< Candidate > &candidates, std::vector< Feature > &features, size_t num_features, float distance)
 Choose candidate features so that they are not bunched together. More...
 

Detailed Description

Represents a modality operating over an image pyramid.

Constructor & Destructor Documentation

§ ~QuantizedPyramid()

virtual cv::linemod::QuantizedPyramid::~QuantizedPyramid ( )
inlinevirtual

Member Function Documentation

§ extractTemplate()

virtual bool cv::linemod::QuantizedPyramid::extractTemplate ( Template templ) const
pure virtual

Extract most discriminant features at current pyramid level to form a new template.

Parameters
[out]templThe new template.

§ pyrDown()

virtual void cv::linemod::QuantizedPyramid::pyrDown ( )
pure virtual

Go to the next pyramid level.

Todo:
Allow pyramid scale factor other than 2

§ quantize()

virtual void cv::linemod::QuantizedPyramid::quantize ( Mat dst) const
pure virtual

Compute quantized image at current pyramid level for online detection.

Parameters
[out]dstThe destination 8-bit image. For each pixel at most one bit is set, representing its classification.

§ selectScatteredFeatures()

static void cv::linemod::QuantizedPyramid::selectScatteredFeatures ( const std::vector< Candidate > &  candidates,
std::vector< Feature > &  features,
size_t  num_features,
float  distance 
)
staticprotected

Choose candidate features so that they are not bunched together.

Parameters
[in]candidatesCandidate features sorted by score.
[out]featuresDestination vector of selected features.
[in]num_featuresNumber of candidates to select.
[in]distanceHint for desired distance between features.

The documentation for this class was generated from the following file: