astro.fits

Class FitsReduction

java.lang.Object

astro.fits.FitsReduction

public class FitsReduction
extends Object

Author:

wooz

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	FitsReduction.Write

Field Summary

Fields

Modifier and Type	Field and Description
private static String	FILTER
private static String	RASTER

Constructor Summary

Constructors

Constructor and Description
FitsReduction()

Method Summary

Modifier and Type	Method and Description
static ArrayLayout	checkLayout(MImageHDU calib, ArrayLayout against)
static float[]	gauss(MImageHDU hdu, double fwhm, double trunc) Uses java image libraries for a fast truncated Gauss convolution filtering.
static Pair<StarDetection,Double>	getClosestToPosition(Point2D at, List<StarDetection> raw, double maxdist) Scans for a StarDetection around the requested position.
static List<StarDetection>	getDetections(MImageHDU reduced, float[] trg, int datamin, int datamax, int searchbox, int thick, int skyframe, double threshold, float[] xst, float[] yst) Tries to detect stars in a reduced image.
private static float[]	getFrame(float[] adu, ArrayLayout layout, int xlf, int yup, int wdt, int hgt, int tck)
static List<StarDetection>	getNeighbours(Point2D at, List<StarDetection> all, double pixdist) Returns all detections within a certain pixel distance to a stated point.
static float[]	read(MImageHDU calib)
static MImageHDU	reduce(MImageHDU raw, double exptime, MImageHDU bias, MImageHDU dark, MImageHDU flat) /** From a raw image we do basic reduction steps by subtracting the bias, scaling the dark to the exposure time and also subtracting it from the raw image and finally flat-fielding.
static float[]	reduceFloat(MImageHDU raw, double exptime, MImageHDU bias, MImageHDU dark, MImageHDU flat) From a raw image we do basic reduction steps by subtracting the bias, scaling the dark to the exposure time and also subtracting it from the raw image and finally flat-fielding.
private static float[]	stripesM(MImageHDU hdu, double fwhm, double trunc, boolean xdir)
static float[]	xstripesM(MImageHDU reduced, double fwhm, double tr) Convolution with a x-boxed Gaussian, thus a gaussian in x and on/off in y.
static float[]	ystripesM(MImageHDU reduced, double fwhm, double tr) Convolution with a y-boxed Gaussian, thus on/off in x and a gaussian in y.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

FILTER

private static final String FILTER

See Also:

Constant Field Values

RASTER

private static final String RASTER

See Also:

Constant Field Values

Constructor Detail

FitsReduction

public FitsReduction()

Method Detail

reduce

public static MImageHDU reduce(MImageHDU raw,
                               double exptime,
                               MImageHDU bias,
                               MImageHDU dark,
                               MImageHDU flat)
                        throws BasicFitsException

/** From a raw image we do basic reduction steps by subtracting the bias, scaling the dark to the exposure time and also subtracting it from the raw image and finally flat-fielding. A step is skipped without exception, if any of the bias/dark/flat frames are missing.

Parameters:

raw - Raw image data

exptime - Exposure time of the frame in seconds

bias - Bias of the camera or null

dark - Dark current of the camera, i.e. normalized to 1 sec exposure or null

flat - Flat-fielding normalization or null

Returns:

The reduced frame (raw-bais-texp*dark)/flat

Throws:

BasicFitsException

checkLayout

public static ArrayLayout checkLayout(MImageHDU calib,
                                      ArrayLayout against)

read

public static float[] read(MImageHDU calib)

reduceFloat

public static float[] reduceFloat(MImageHDU raw,
                                  double exptime,
                                  MImageHDU bias,
                                  MImageHDU dark,
                                  MImageHDU flat)
                           throws BasicFitsException

From a raw image we do basic reduction steps by subtracting the bias, scaling the dark to the exposure time and also subtracting it from the raw image and finally flat-fielding. A step is skipped without exception, if any of the bias/dark/flat frames are missing, i.e., it is null.

Parameters:

raw - Raw image data

exptime - Exposure time of the frame in seconds

bias - Bias/Null of the camera

dark - Dark/Null current of the camera, i.e. normalized to 1 sec exposure

flat - Flat-fielding normalization/Null

Returns:

The reduced image as a float array, geometry as original.

Throws:

BasicFitsException

gauss

public static float[] gauss(MImageHDU hdu,
                            double fwhm,
                            double trunc)
                     throws BasicFitsException

Uses java image libraries for a fast truncated Gauss convolution filtering. The built-in data type of unsigned short makes distinction to negative pixel values (resulting from lowering the Gaussian) impossible, thus only the positive half of the short range is used (ADUs above are limited to Short.MAX_VALUE). This still may lead to problems if the Gaussian filtering FWHM is totally off of the stellar images expected, but the speed gain justifies even this shortcoming.

Parameters:

hdu - The (reduced) image

fwhm - pixel-fwhm of the Gaussian convolution.

trunc - The truncation radius in FWHM.

Returns:

Throws:

BasicFitsException

FitsException

xstripesM

public static float[] xstripesM(MImageHDU reduced,
                                double fwhm,
                                double tr)
                         throws BasicFitsException

Convolution with a x-boxed Gaussian, thus a gaussian in x and on/off in y.

Parameters:

reduced -

fwhm - pixel-fwhm of the Gaussian convolution.

tr - The truncation radius in FWHM.

Returns:

Throws:

FitsException

BasicFitsException

ystripesM

public static float[] ystripesM(MImageHDU reduced,
                                double fwhm,
                                double tr)
                         throws BasicFitsException

Convolution with a y-boxed Gaussian, thus on/off in x and a gaussian in y.

Parameters:

reduced -

fwhm - pixel-fwhm of the Gaussian convolution.

tr - The truncation radius in FWHM.

Returns:

Throws:

FitsException

BasicFitsException

stripesM

private static float[] stripesM(MImageHDU hdu,
                                double fwhm,
                                double trunc,
                                boolean xdir)
                         throws BasicFitsException

Throws:

BasicFitsException

getDetections

public static List<StarDetection> getDetections(MImageHDU reduced,
                                                float[] trg,
                                                int datamin,
                                                int datamax,
                                                int searchbox,
                                                int thick,
                                                int skyframe,
                                                double threshold,
                                                float[] xst,
                                                float[] yst)
                                         throws BasicFitsException

Tries to detect stars in a reduced image. For better detection results, a Gaussian convolved image should be provided of identical layout than the reduced image. For shutterless images, a pair of x/y enhanced stripes can be useful to detect and discard read-out stripes.

Parameters:

reduced - The bias-subtracted, dark corrected and flat-fielded image

gauss - The array of a truncated Gaussian filtered image

datamin - Minimum ADU to account for a star signal.

datamax - Maximum ADU to account for a star signal.

searchbox - All detected pixel within this box belong to a star. If too small, a star might show up as two detections. If too big, binaries may show up only as a single detection.

thick - Edge region with artifacts from Gauss convolution, pixel.

skyframe - Thickness around searchbox for background determination

threshold - Signal is higher than this times background σ.

xstripes - Readout stripes in x-direction (rectangular Gaussian)

ystripes - Readout stripes in y-direction (rectangular Gaussian)

Returns:

A list of star detection, including spurious & stripes, ordered to magnitude

Throws:

BasicFitsException

getFrame

private static float[] getFrame(float[] adu,
                                ArrayLayout layout,
                                int xlf,
                                int yup,
                                int wdt,
                                int hgt,
                                int tck)
                         throws BasicFitsException

Throws:

BasicFitsException

getClosestToPosition

public static Pair<StarDetection,Double> getClosestToPosition(Point2D at,
                                                              List<StarDetection> raw,
                                                              double maxdist)

Scans for a StarDetection around the requested position. Returns the closest object plus distance or null if no object is within reach. Position refers to possible binned and windowed read-out, i.e., as extracted by #getDetectedStars.

Parameters:

at - Search closest to this position as extracted from CCD window.

raw - The list of stars detected.

maxdist - Maximum pixel distance. Return any if zero or negative.

Returns:

The closest detection and its distance in pixel. If no target is found, a null-pair is returned rather than a null First

getNeighbours

public static List<StarDetection> getNeighbours(Point2D at,
                                                List<StarDetection> all,
                                                double pixdist)

Returns all detections within a certain pixel distance to a stated point. Never returns null, at most an empty list.

Parameters:

at - Search closest to this position as extracted from CCD window.

all - All stars detected in frame.

pixdist - Maximum pixel distance to be considered a neighbour

Returns: