but_plane_detector::Regions Class Reference

#include <filtering.h>

Public Member Functions
bool	computeStatistics (float threshold)
void	independentTileRegions (cv::Mat &src, const sensor_msgs::CameraInfoConstPtr &cam_info, float minimumPlaneDistance=0.01, float minimumTriDistance=0.02, float minimumFloodDistance=0.05, int tileSize=30, int smallestTriangleArea=100, int minRegionSize=10, int minCandidates=10)
	Regions (Normals *normals=NULL)
void	watershedRegions (cv::Mat &src, const sensor_msgs::CameraInfoConstPtr &cam_info, int type=WatershedType::Combined, int seedThreshold=1, float alpha=0.0, float beta=0.0, float gamma=0.0)
	~Regions ()
Static Public Member Functions
static void	gradientDepthDifference (cv::Mat &src, cv::Mat &dst, int size, float differenceThreshold)
static void	gradientNormalDifference (cv::Mat &src, cv::Mat &dst, int size, float differenceThreshold)
static void	gradientPlanePredictor (cv::Mat &src, cv::Mat &dst, float differenceThreshold=0.02, int size=2, int sobelSize=5)
Public Attributes
Normals *	m_normals
cv::Mat	m_planes
cv::Mat	m_regionMatrix
cv::Mat	m_stddeviation
Private Member Functions
void	fillEverything (Plane< float > &plane, cv::Mat &depth, cv::Mat &mask, cv::Mat &points, int i_tile, int j_tile, int tileSize, int index, float min_planeDistance=0.05)
int	floodFillTile (cv::Mat &tile, cv::Mat &tileMask, cv::Mat &pointMask, cv::Vec3f *points, int index, Plane< float > &plane, int ioffset, int joffset, float min_planeDistance=0.02)
void	getLeastSquaresAndBorder (Plane< float > &plane, cv::Mat &tile, cv::Mat &tileMask, int ioffset, int joffset, int index)

Detailed Description

Class providing an interface to depth image segmentation

Definition at line 85 of file filtering.h.

Constructor & Destructor Documentation

but_plane_detector::Regions::Regions ( Normals * normals = NULL )

A class providing an interface to depth image segmenter

Parameters:

normals

Normals object with precomputed real points and normals - not necessary - only if you do not want to compute it in this module (for speed reasons)

See also:: Normals()

Definition at line 55 of file filtering.cpp.

but_plane_detector::Regions::~Regions ( )

Destructor - obviously, it destructs this class

Definition at line 63 of file filtering.cpp.

Member Function Documentation

bool but_plane_detector::Regions::computeStatistics ( float threshold )

Method computes from m_regionMatrix statistical information (fills m_planes and m_stddeviation matrices)

Parameters:

threshold

Not used now (TODO)

Definition at line 73 of file filtering.cpp.

void but_plane_detector::Regions::fillEverything	(	Plane< float > &	plane,
		cv::Mat &	depth,
		cv::Mat &	mask,
		cv::Mat &	points,
		int	i_tile,
		int	j_tile,
		int	tileSize,
		int	index,
		float	min_planeDistance = `0.05`
	)			`[private]`

Auxiliary function for filling entire image from mask seed (until threshold from plane is met)

Parameters:

	plane	Plane equation
	depth	Depth image
	mask	Region mask
	points	Vector of seed points
	i_tile	Row tile offset
	j_tile	Column tile offset
	tileSize	Size of tile
	index	Index of filled region
	min_planeDistance	Maximal distance from plane to be filled

int but_plane_detector::Regions::floodFillTile	(	cv::Mat &	tile,
		cv::Mat &	tileMask,
		cv::Mat &	pointMask,
		cv::Vec3f *	points,
		int	index,
		Plane< float > &	plane,
		int	ioffset,
		int	joffset,
		float	min_planeDistance = `0.02`
	)			`[private]`

Auxiliary method which flood fills a tile untill min_planeDistance threshold difference is met

Parameters:

	tile	Tile subimage of depth data
	tileMask	Mask subimage
	pointMask	Point mask subimage
	points	Vector of tile points
	index	Index of filling region
	plane	Equation of filled plane
	ioffset	Row offset of tile
	joffset	Column offset of tile
	min_planeDistance	Minimal distance from plane to fill

void but_plane_detector::Regions::getLeastSquaresAndBorder	(	Plane< float > &	plane,
		cv::Mat &	tile,
		cv::Mat &	tileMask,
		int	ioffset,
		int	joffset,
		int	index
	)			`[private]`

Auxiliary function extracts current region's plane using LSQ

Parameters:

	plane	Extracted plane
	tile	Tile subimage of depth data
	tileMask	Mask subimage
	ioffset	Row offset of tile
	joffset	Column offset of tile
	index	Index of filling region

static void but_plane_detector::Regions::gradientDepthDifference	(	cv::Mat &	src,
		cv::Mat &	dst,
		int	size,
		float	differenceThreshold
	)			`[static]`

Method converts an input image into gradient image using comparison of depth data

Parameters:

	src	Input CV_16UC depth matrix (raw input from kinect)
	dst	Final gradient image
	size	Size of neigborhood (number of maximal pixel distance)
	differenceThreshold	Difference from center depth to mark a pixel as anomal

static void but_plane_detector::Regions::gradientNormalDifference	(	cv::Mat &	src,
		cv::Mat &	dst,
		int	size,
		float	differenceThreshold
	)			`[static]`

Method converts an input image into gradient image using comparison of normals

Parameters:

	src	Input CV_16UC depth matrix (raw input from kinect)
	dst	Final gradient image
	size	Size of neigborhood (number of maximal pixel distance)
	differenceThreshold	Difference from center normal to mark a pixel as anomal

static void but_plane_detector::Regions::gradientPlanePredictor	(	cv::Mat &	src,
		cv::Mat &	dst,
		float	differenceThreshold = `0.02`,
		int	size = `2`,
		int	sobelSize = `5`
	)			`[static]`

Method converts an input image into gradient image using expected plane prediction. Pixel value is number of pixels around which are not in difference threshold from expected plane.

Parameters:

	src	Input CV_16UC depth matrix (raw input from kinect)
	dst	Final gradient image
	differenceThreshold	Difference from expected plane threshold to mark a pixel as anomal
	size	Size of neigborhood (number of maximal pixel distance)
	sobelSize	Size of sobel kernel used for gradient image computation

void but_plane_detector::Regions::independentTileRegions	(	cv::Mat &	src,
		const sensor_msgs::CameraInfoConstPtr &	cam_info,
		float	minimumPlaneDistance = `0.01`,
		float	minimumTriDistance = `0.02`,
		float	minimumFloodDistance = `0.05`,
		int	tileSize = `30`,
		int	smallestTriangleArea = `100`,
		int	minRegionSize = `10`,
		int	minCandidates = `10`
	)

Method segments a depth image using own tile plane searching RANSAC method.

Parameters:

	src	Input CV_16UC depth matrix (raw input from kinect)
	cam_info	Camera info message (ROS)
	minimumPlaneDistance	Minimal distance threshold from triangle plane in whole tile (in meters)
	minimumTriDistance	Minimal distance threshold from triangle plane in triangle flood fill (in meters)
	minimumFloodDistance	Minimal distance threshold from triangle plane in whole image (in meters)
	tileSize	Tile size
	smallestTriangleArea	Smalest randomly found triangle to be considered for plane computation (in pixels^2)
	minRegionSize	Minimal region size to be considered
	minCandidates	Minimal number of pixels in tile to start a RANSAC

void but_plane_detector::Regions::watershedRegions	(	cv::Mat &	src,
		const sensor_msgs::CameraInfoConstPtr &	cam_info,
		int	type = `WatershedType::Combined`,
		int	seedThreshold = `1`,
		float	alpha = `0.0`,
		float	beta = `0.0`,
		float	gamma = `0.0`
	)

Method which segments a depth image with watershed algorithm. There is several approaches, so please be aware of input parameters

Parameters:

	src	Input CV_16UC depth matrix (raw input from kinect)
	cam_info	Camera info message (ROS)
	type	Type of watershed input - one of WatershedType constants

See also:: WatershedType

Parameters:

	seedThreshold	Threshold which specifies number of anomal neighbors for selecting this pixel as seed for watershed
	alpha	It is always size of neigborhood which is considered for searching the image in watershed preprocessing. If 0.0, default is selected.
	beta	When simple preprocessing is selected (depth, normal, predictor), it is always a threshold for selecting anomal neighbors (in depth is in milimeters, normal is in radians ). If combined type is selected, it represents a depth difference.
	gamma	It is used only in combined and predictor. In combined, it specifiesd normal difference threshold, in predictor a sobel size.

See also:: Normals()

Member Data Documentation

Normals* but_plane_detector::Regions::m_normals

Instance of Normals object - precomputed real point positions and normals. Standard is null if not specified otherwise in constructor or in Region method Freed in destructor!

Definition at line 188 of file filtering.h.

cv::Mat but_plane_detector::Regions::m_planes

Computed planes matrix (Vec4f type, each component represents component of computed region normal) Filled after computeStatistics method