Tries to find good parameters (which are threshold1, threshold2, blurwidth) for a given image and mask. More...

Public Member Functions
	AreaParameterChooserWorker (ros::NodeHandle &node, cv::Mat &image, cv::Mat &mask, const std::map< std::string, std::map< std::string, double > > &targetValue)
void	CallculateParameters (int numResults, int steps, std::vector< face_contour_detector::autoselect_result > &result)
Private Member Functions
void	ApplyFilters (const std::vector< double > &params, std::vector< face_contour_detector::filter_area_result > &results)
double	CostFunction (const std::map< std::string, double > &result, const std::map< std::string, double > &targetValue)
void	GetAreas ()
	uses the area service to load the areas
void	ParamsFunction (const std::vector< double > &params, std::map< std::string, double > &result)
	callbacks for the AutoParamterExplorer
void	PushFilterSetup (std::map< std::string, std::vector< std::map< std::string, double > > > &params, int numResults, std::vector< face_contour_detector::autoselect_result > &result)
void	SendStatusMessage (const std::string &message)
Private Attributes
std::vector < face_contour_detector::image_area >	areas
cv::Mat	bwimage
std::vector < face_contour_detector::image_area > ::iterator	currArea
cv::Mat &	image
face_contour_detector::KdTree < double, std::map < std::string, double > >	kdTree
cv::Mat &	mask
ros::NodeHandle &	node
std::map< std::string, std::map< std::string, double > >	targetValue

Detailed Description

Tries to find good parameters (which are threshold1, threshold2, blurwidth) for a given image and mask.

Constructor & Destructor Documentation

face_contour_detector::AreaParameterChooserWorker::AreaParameterChooserWorker	(	ros::NodeHandle &	node,
		cv::Mat &	image,
		cv::Mat &	mask,
		const std::map< std::string, std::map< std::string, double > > &	targetValue
	)

Constructor

Parameters:

node	ros node to call the area service
image	the image
mask	the mask image, this is only used to get the areas from the area service
targetValue	contains the target values for the different areas. In this case only the "graphmass" can be set for each area

void face_contour_detector::AreaParameterChooserWorker::ApplyFilters	(	const std::vector< double > &	params,
		std::vector< face_contour_detector::filter_area_result > &	results
	)		`[private]`

applies the filters and returns the result properties (currently only returns the graphmass) Note: this not beeing used currently

Parameters:

params	the parameters (threshold1, threshold2, blurwidth)
results	the results

void face_contour_detector::AreaParameterChooserWorker::CallculateParameters	(	int	numResults,
		int	steps,
		std::vector< face_contour_detector::autoselect_result > &	result
	)

Starts the search for good values for Canny->threshold1, Canny->threshold2 and GaussianBlur->width/height

Parameters:

numResults	the number of results that should be returned
steps	the maximum number of subdivision steps that should be taken by the AutoParameterExplorer
result	the filter setup results, this can be send to the display_on_gui service directly

double face_contour_detector::AreaParameterChooserWorker::CostFunction	(	const std::map< std::string, double > &	result,
		const std::map< std::string, double > &	targetValue
	)		`[private]`

void face_contour_detector::AreaParameterChooserWorker::GetAreas ( ) [private]

uses the area service to load the areas

void face_contour_detector::AreaParameterChooserWorker::ParamsFunction	(	const std::vector< double > &	params,
		std::map< std::string, double > &	result
	)		`[private]`

callbacks for the AutoParamterExplorer

void face_contour_detector::AreaParameterChooserWorker::PushFilterSetup	(	std::map< std::string, std::vector< std::map< std::string, double > > > &	params,
		int	numResults,
		std::vector< face_contour_detector::autoselect_result > &	result
	)		`[private]`