OpenCV  3.2.0-dev
Open Source Computer Vision

#include "cudalegacy.hpp"

Collaboration diagram for cv::cuda::FGDParams:

Public Member Functions

 FGDParams ()
 default Params More...
 

Public Attributes

float alpha1
 How quickly we forget old background pixel values seen. More...
 
float alpha2
 "Controls speed of feature learning". More...
 
float alpha3
 Alternate to alpha2, used (e.g.) for quicker initial convergence. More...
 
float delta
 Affects color and color co-occurrence quantization, typically set to 2. More...
 
bool is_obj_without_holes
 If TRUE we ignore holes within foreground blobs. More...
 
int Lc
 Quantized levels per 'color' component. More...
 
int Lcc
 Quantized levels per 'color co-occurrence' component. More...
 
float minArea
 Discard foreground blobs whose bounding box is smaller than this threshold. More...
 
int N1c
 Number of color vectors used to model normal background color variation at a given pixel. More...
 
int N1cc
 Number of color co-occurrence vectors used to model normal background color variation at a given pixel. More...
 
int N2c
 Number of color vectors retained at given pixel. More...
 
int N2cc
 Number of color co-occurrence vectors retained at given pixel. More...
 
int perform_morphing
 Number of erode-dilate-erode foreground-blob cleanup iterations. More...
 
float T
 A percentage value which determines when new features can be recognized as new background. More...
 

Constructor & Destructor Documentation

cv::cuda::FGDParams::FGDParams ( )

default Params

Member Data Documentation

float cv::cuda::FGDParams::alpha1

How quickly we forget old background pixel values seen.

Typically set to 0.1.

float cv::cuda::FGDParams::alpha2

"Controls speed of feature learning".

Depends on T. Typical value circa 0.005.

float cv::cuda::FGDParams::alpha3

Alternate to alpha2, used (e.g.) for quicker initial convergence.

Typical value 0.1.

float cv::cuda::FGDParams::delta

Affects color and color co-occurrence quantization, typically set to 2.

bool cv::cuda::FGDParams::is_obj_without_holes

If TRUE we ignore holes within foreground blobs.

Defaults to TRUE.

int cv::cuda::FGDParams::Lc

Quantized levels per 'color' component.

Power of two, typically 32, 64 or 128.

int cv::cuda::FGDParams::Lcc

Quantized levels per 'color co-occurrence' component.

Power of two, typically 16, 32 or 64.

float cv::cuda::FGDParams::minArea

Discard foreground blobs whose bounding box is smaller than this threshold.

int cv::cuda::FGDParams::N1c

Number of color vectors used to model normal background color variation at a given pixel.

int cv::cuda::FGDParams::N1cc

Number of color co-occurrence vectors used to model normal background color variation at a given pixel.

int cv::cuda::FGDParams::N2c

Number of color vectors retained at given pixel.

Must be > N1c, typically ~ 5/3 of N1c. Used to allow the first N1c vectors to adapt over time to changing background.

int cv::cuda::FGDParams::N2cc

Number of color co-occurrence vectors retained at given pixel.

Must be > N1cc, typically ~ 5/3 of N1cc. Used to allow the first N1cc vectors to adapt over time to changing background.

int cv::cuda::FGDParams::perform_morphing

Number of erode-dilate-erode foreground-blob cleanup iterations.

These erase one-pixel junk blobs and merge almost-touching blobs. Default value is 1.

float cv::cuda::FGDParams::T

A percentage value which determines when new features can be recognized as new background.

(Typically 0.9).


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