analyseMatrix Class Reference

#include <analysematrix.h>

List of all members.

Public Member Functions

• analyseMatrix ()
• ~analyseMatrix ()
• void filterBy2x2 (int filter[2][2], int col, int row, double matrix[20][22], double res[20][22], int mask[20][22])
• void displayMatrix (FILE *fout, double matrix[20][22], const char *title)
• void displayMatrix (FILE *fout, int matrix[20][22], const char *title, int norm=0, int value=0, const char *tag=0)
• int displayPixelList (FILE *fout, int matrix[20][22])
• void classifyByThreshold (double thresh, double matrix[20][22], int res[20][22], int up=1)
• void setMatrix (double value, double matrix[20][22])
• void setMatrix (int value, int matrix[20][22])
• void multMatrix (double value, double matrix[20][22])
• void multMask (int value, int mask[20][22])
• void divMatrix (double res[20][22], double matrix[20][22])
• void divMatrix (double res[20][22], int matrix[20][22])
• void meanMatrix (double matrix[20][22], double *mean, double *var, int *n, int mask[20][22])
• void meanMatrix (double matrix[20][22], double *mean, double *var, int *n, int mask[20][22], int err[20][22])
• void addMask (int res[20][22], int mask[20][22])
• void addMatrix (double res[20][22], double matrix[20][22])
• void copyMatrix (double res[20][22], double matrix[20][22])
• void orMask (int res[20][22], int mask[20][22])
• void invertMask (int mask[20][22])
• int nMask (int mask[20][22])
• void copyMask (int res[20][22], int mask[20][22])
• void maxMask (int max, int mask[20][22])
• void minMask (int min, int mask[20][22])
• void findHighNoise1 (FILE *fout, double matrix[20][22], int res[20][22], int err[20][22])
• void findHighNoise2 (FILE *fout, double matrix[20][22], int res[20][22], int err[20][22])
• void clearNoisyPixel ()
• void clearFaultyPixel ()
• void addNoisyPixel (int mask[20][22])
• void findHighNoise (FILE *fout, double matrix[20][22])
• void meanOfNormalPixel (double matrix[20][22], double *mean, double *var, int *n)
• void meanOfNoisyPixel (double matrix[20][22], double *mean, double *var, int *n)
• int nNoisyPixel ()
• void displayNoisyPixelList (FILE *fout)

Public Attributes

• int noisyPixel [20][22]
• int noisyPixelLast [20][22]
• int faultyPixel [20][22]
• int noisyRuns
• double noisyNVariances
• int noisyMinVariance
• double noisyNeighborDiff
• double noisyAbsThresh
• double noisyLowThresh
• int debug

---

Detailed Description

Analyse the variance matrix of a telescope All functions are on matrices of 20x22 elements.

Definition at line 22 of file analysematrix.h.

---

Constructor & Destructor Documentation

analyseMatrix::analyseMatrix

(

)

analyseMatrix::~analyseMatrix

(

)

---

Member Function Documentation

void analyseMatrix::addMask	(	int	res[20][22],
		int	mask[20][22]
	)

Add two masks.

void analyseMatrix::addMatrix	(	double	res[20][22],
		double	matrix[20][22]
	)

Add two matrixes.

void analyseMatrix::addNoisyPixel

(

int

mask[20][22]

)

Add a new noisy pixel mask.

void analyseMatrix::classifyByThreshold	(	double	thresh,
		double	matrix[20][22],
		int	res[20][22],
		int	up = 1
	)

Classify by using a constant threshold matrix, res and mask are 20x22 pixel.

The up parameter controls whether the values higher or lower of the threshold should be taken.

void analyseMatrix::clearFaultyPixel

(

)

Clear faulty pixel matrix.

void analyseMatrix::clearNoisyPixel

(

)

Clear noisy pixel matrix.

void analyseMatrix::copyMask	(	int	res[20][22],
		int	mask[20][22]
	)

Copy the mask to res.

void analyseMatrix::copyMatrix	(	double	res[20][22],
		double	matrix[20][22]
	)

Copy matrix to res.

void analyseMatrix::displayMatrix	(	FILE *	fout,
		double	matrix[20][22],
		const char *	title
	)

Display a matrix of 20x22 elements.

void analyseMatrix::displayMatrix	(	FILE *	fout,
		int	matrix[20][22],
		const char *	title,
		int	norm = 0,
		int	value = 0,
		const char *	tag = 0
	)

Display a matrix of 20x22 integer elements.

With the argument norm > 0, relative values in percent a the normation given will be displayed. A special value can be replaced by a tag.

void analyseMatrix::displayNoisyPixelList

(

FILE *

fout

)

Display the list of noisy pixel in the last analysis.

int analyseMatrix::displayPixelList	(	FILE *	fout,
		int	matrix[20][22]
	)

Display list of coordinates of all pixel in the mask Returns the number of pixel in the mask.

void analyseMatrix::divMatrix	(	double	res[20][22],
		double	matrix[20][22]
	)

Devide by matrix.

void analyseMatrix::divMatrix	(	double	res[20][22],
		int	matrix[20][22]
	)

Devide by matrix.

void analyseMatrix::filterBy2x2	(	int	filter[2][2],
		int	col,
		int	row,
		double	matrix[20][22],
		double	res[20][22],
		int	mask[20][22]
	)

Filter a matrix using a free 2x2 matrix.

With mask pixel can be excluded from the calculation matrix, res and mask are 20x22 pixel Parameters (col/row) determine the center pixel of the filter. Possible values are col, row = {0|1}.

void analyseMatrix::findHighNoise	(	FILE *	fout,
		double	matrix[20][22]
	)

Analyse matrix and store result in the local arrays.

void analyseMatrix::findHighNoise1	(	FILE *	fout,
		double	matrix[20][22],
		int	res[20][22],
		int	err[20][22]
	)

Automatically find pixel with high noise in the variance matrix.

Algorithm:

• Calculate statistics of pixel distribution
• Take threshold proportional to N times the distribution variance. Consider a minumum threshold.
• Repeat the algorithm taking only non noisy pixel into account.

Parameter of the alorithm:

• Factor N of the variances [-]
• Minimum threshold [ADC Counts]

void analyseMatrix::findHighNoise2	(	FILE *	fout,
		double	matrix[20][22],
		int	res[20][22],
		int	err[20][22]
	)

Automatically find pixel with high noise in the variance matrix.

Algorithm:

• Find all negative values and exclude from further analsysis
• Run a 2x2 edge filters over the pixel matrix.
• Use a threshold on the difference matrix.
• Repeat this rotating the 2x2 filter to consider all orientations. and overlap the resulting pixel mask.
• Calculate the distribution mean for low and high noise pixel
• Use a threshold on the orignal distribution. The thresold is choosen to mean high noise value minus half of the difference image threshold. Add the results.

Parameter of the algorithm:

• Threshold in the difference matrixes

void analyseMatrix::invertMask

(

int

mask[20][22]

)

Invert the values of the mask.

void analyseMatrix::maxMask	(	int	max,
		int	mask[20][22]
	)

Maximum of constant and mask values.

void analyseMatrix::meanMatrix	(	double	matrix[20][22],
		double *	mean,
		double *	var,
		int *	n,
		int	mask[20][22]
	)

Calculate the mean value of a matrix.

Mask can be used to exclude some pixel from mean value

void analyseMatrix::meanMatrix	(	double	matrix[20][22],
		double *	mean,
		double *	var,
		int *	n,
		int	mask[20][22],
		int	err[20][22]
	)

void analyseMatrix::meanOfNoisyPixel	(	double	matrix[20][22],
		double *	mean,
		double *	var,
		int *	n
	)

Mean of noisy pixel Faulty pixel are not considered.

void analyseMatrix::meanOfNormalPixel	(	double	matrix[20][22],
		double *	mean,
		double *	var,
		int *	n
	)

Mean of normal pixel.

Faulty pixel are not considered

void analyseMatrix::minMask	(	int	min,
		int	mask[20][22]
	)

Minimum of constant and mask values.

void analyseMatrix::multMask	(	int	value,
		int	mask[20][22]
	)

Multiply mask by constant.

void analyseMatrix::multMatrix	(	double	value,
		double	matrix[20][22]
	)

Multiply matrix by constant.

int analyseMatrix::nMask

(

int

mask[20][22]

)

Count the pixel in the mask.

int analyseMatrix::nNoisyPixel

(

)

Number of noisy pixel.

void analyseMatrix::orMask	(	int	res[20][22],
		int	mask[20][22]
	)

Merge two mask using a logical or.

void analyseMatrix::setMatrix	(	double	value,
		double	matrix[20][22]
	)

Set matrix to constant value.

void analyseMatrix::setMatrix	(	int	value,
		int	matrix[20][22]
	)

---

Member Data Documentation

int analyseMatrix::debug

Debug flag.

Definition at line 220 of file analysematrix.h.

int analyseMatrix::faultyPixel[20][22]

Matrix of faulty pixel.

Every time the statistical evaluation finds pixel with neg variances they were marked in the faulty pixel matrix and excluded from further calculations

Definition at line 193 of file analysematrix.h.

double analyseMatrix::noisyAbsThresh

Absolut threshold used to find noisy pixel relativ to the distribution mean.

Definition at line 212 of file analysematrix.h.

double analyseMatrix::noisyLowThresh

Find extreme low noise pixel?! These pixel will be added to the error pixel.

Definition at line 216 of file analysematrix.h.

int analyseMatrix::noisyMinVariance

Tolerated varinance deviation.

Definition at line 206 of file analysematrix.h.

double analyseMatrix::noisyNeighborDiff

Difference between neiboring pixel relativ to the distribution mean.

Definition at line 209 of file analysematrix.h.

double analyseMatrix::noisyNVariances

Tolerated variance changes proportional to variance distribution over the telescope.

Definition at line 202 of file analysematrix.h.

int analyseMatrix::noisyPixel[20][22]

Matrix of noisy pixel.

Every time the statistical evaluation finds pixel with high noise they were marked in the noisy pixel matrix and excluded from further calculations

Definition at line 185 of file analysematrix.h.

int analyseMatrix::noisyPixelLast[20][22]

Last noisy pixel mask calculated by findNoisyPixel()

Definition at line 188 of file analysematrix.h.

int analyseMatrix::noisyRuns

Count the number of analysis runs per telescope.

Definition at line 196 of file analysematrix.h.

---

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

• /home/kopmann/git-mirror/fdhwlib/fdhwlib/akutil/analysematrix.h