PUBLIC

RegBoundaryC::RegBoundaryC(void) RegBoundaryC::RegBoundaryC(const ImageRectangleC &,const NumLabelC &) RegBoundaryC::RegBoundaryC(const RegBoundaryC &) RegBoundaryC::RegBoundaryC(RegBoundaryBC &) RegBoundaryC::Copy(void) const RegBoundaryC::operator=(const RegBoundaryC &) RegBoundaryC::AddPixel(const PixelC &,const ByteImageC &) RegBoundaryC::AverageContrast(void) const RegBoundaryC::GradientContrast(void) const RegBoundaryC::Compactness(void) const RegBoundaryC::GetNextPixel(void) RegBoundaryC::AddRegionToImage(ImageC &) const RegBoundaryC::AddOverlayToImage(RGBImageC &,const ByteRGBValueC &) const RegBoundaryC::GetLabel(void) const RegBoundaryC::SetLabel(const NumLabelC &) RegBoundaryC::GetCluster(void) const RegBoundaryC::SetCluster(const NumLabelC &) RegBoundaryC::GetRegionSize(void) const RegBoundaryC::GetMeanPos(void) const RegBoundaryC::GetMotion(void) const RegBoundaryC::SetMotion(Vector2dC) RegBoundaryC::GetMoments(void) const RegBoundaryC::SetPreviousRegion(RegBoundaryC &) RegBoundaryC::PreviousRegion(void) RegBoundaryC::IsValidObject(void) const RegBoundaryC::GetNeighbourhood(void) const

RegBoundaryC
Stores a region boundary as well as statistcs
include	"amma/DERA/RegBoundary.hh"
User Level:	Default
Library:	DERARegTrack
Example:	exIndex.cc
Section:	Applications.Image Processing.DERA Small Target Detection
In Scope:	std

Comments:
This is a handle class to RegBoundaryBC which actually does the work, ie it is a BIG object. This approach means that the regions can easily be stored in data structures and have muliple handles to them without having to resort to pointers.

Used by RegOrderedGrowC to represent intermediate stages in the region growing process. It stores a labelled image of the region as well as statistics of the region and boundary. The statistics are updated incrementally when a pixel is added to the region.

RegBoundaryC()

Default constructor

ind	Index of types of intererest.

RegBoundaryC(const ImageRectangleC & rect,const NumLabelC & label)

Constructor

ind	Index of types of intererest.

This sets up a region ready for growing in an image of size rect and with the given label value.

RegBoundaryC(const RegBoundaryC & oth)

Copy constructor

ind	Index of types of intererest.

RegBoundaryC(RegBoundaryBC & oth)

Constructs from implementation class

ind	Index of types of intererest.

RegBoundaryC Copy() const

Makes a deep copy

ind	Index of types of intererest.

RegBoundaryC & operator=(const RegBoundaryC & oth)

Assignment operator

ind	Index of types of intererest.

void AddPixel(const PixelC & pos,const ByteImageC & scene)

Adds a pixel to the region

ind	Index of types of intererest.

This involves the following steps: maintain a bounding box of the region, update boundaries statistics incrementally, add point to current boundary, update internal boundary.

RealT AverageContrast() const

Difference between region and current boundary average intensities

ind	Index of types of intererest.

RealT GradientContrast() const

Difference between internal and current boundaries average intensities

ind	Index of types of intererest.

RealT Compactness() const

Measures geometrical compactness of the region

ind	Index of types of intererest.

This is the ratio of the perimeter to the square root of the area of the region.

const PixelC GetNextPixel()

Finds next pixel from the current boundary

ind	Index of types of intererest.

Uses a priority queue to do this. Points are added to the queue in AddPixel but it is removed here.

void AddRegionToImage(ImageC<NumLabelC> & image) const

Inserts labelled region into image

ind	Index of types of intererest.

The region is labelled with the value specified in the constructor. The boundary is labelled with the same value with bit 16 set.

void AddOverlayToImage(RGBImageC & scene,const ByteRGBValueC & colour) const

Inserts overlay of region into image

ind	Index of types of intererest.

The internal boundary of the region is labelled in the scene image as pixels with the specified colour.

NumLabelC GetLabel() const

Access to the label of the region

ind	Index of types of intererest.

void SetLabel(const NumLabelC & label)

Specify a label value for the region

ind	Index of types of intererest.

NumLabelC GetCluster() const

Access to the motion cluster of the region

ind	Index of types of intererest.

void SetCluster(const NumLabelC & label)

Specify a motion cluster value for the region

ind	Index of types of intererest.

UIntT GetRegionSize() const

The number of pixels in the region

ind	Index of types of intererest.

Vector2dC GetMeanPos() const

Centre of mass of the region

ind	Index of types of intererest.

Vector2dC GetMotion() const

Motion vector of the region

ind	Index of types of intererest.

void SetMotion(Vector2dC motion)

Specify the motion vector of the region

ind	Index of types of intererest.

Matrix2d2C GetMoments() const

Gets 2nd order moments of region as a covariance matrix

ind	Index of types of intererest.

void SetPreviousRegion(RegBoundaryC & previous)

Sets a handle to previous region

ind	Index of types of intererest.

This is used to track the history of the region. At each point in time there is a handle to the region at the previous point in time. This is set during motion tracking.

RegBoundaryC PreviousRegion()

Gets handle to same region in previous image

ind	Index of types of intererest.

BooleanT IsValidObject() const

Checks that handle to previous region is valid

ind	Index of types of intererest.

If there was no correspondance found then the handle has not been set to point to another region and is therefore not valid.

IndexNeigh2dC GetNeighbourhood() const

Returns neighbourhood list of pixels in the region

ind	Index of types of intererest.