IMAGE module

image is the pyvista module that implements some general image processing including basic image arithmetic, VISTA-style, and VISTA-style boxes

image functions:

class pyvista.image.BOX(n=None, nr=None, nc=None, sr=1, sc=1, cr=None, cc=None, xr=None, yr=None)

Defines BOX class

set(xmin, xmax, ymin, ymax)

Resets limits of a box

Parameters

• xmin – lower x value

• xmax – higher x value

• ymin – lower y value

• ymax – higher xyvalue

nrow()

Returns number of rows in a box

Returns :

number of rows

ncol()

Returns number of columns in a box

Returns :

number of columns

show(header=True)

Prints box limits

mean(data)

Returns mean of data in box

Args :

data : input data (Data or np.array)

Returns

mean of data in box

stdev(data)

Returns standard deviation of data in box

Args :

data : input data (Data or np.array)

Returns

standard deviation of data in box

iqr(data)

Returns IQR data in box

Args :

data : input data (Data or np.array)

Returns

IQR of data in box

max(data)

Returns maximum of data in box

Args :

data : input data (Data or np.array)

Returns

maximum of data in box

min(data)

Returns minimum of data in box

Args :

data : input data (Data or np.array)

Returns

minimum of data in box

median(data)

Returns median of data in box

Args :

data : input data (Data or np.array)

Returns

median of data in box

setval(data, val)

Sets data in box to specified value

setbit(data, val)

Sets bit of data in box to specified value

getval(data)

Returns data in box

Args :

data : input data (Data or np.array)

Returns

data in box

pyvista.image.abx(data, box)

Returns dictionary with image statistics in box.

Args :

data : input data (Data or np.array) box : pyvista BOX

Returns :

dictionary with image statistics : ‘mean’, ‘stdev’, ‘min’, ‘max’, ‘peakx’, ‘peaky’

pyvista.image.gauss2d(X, amp, x0, y0, a, b, c, back, binned=False)

Evaluate Gaussian 2D function

Form: amp*exp(-a(x-x0)**2 - b(x-x0)*(y-y0) - c(y-y0)**2) + const

Parameters

• X (arraylike [2,npts]) – x,y positions to evaluate at

• amp (float) – coefficients of function

• x0 (float) – coefficients of function

• y0 (float) – coefficients of function

• a (float) – coefficients of function

• b (float) – coefficients of function

• c (float) – coefficients of function

• back (float) – coefficients of function

pyvista.image.gh2d(X, gpars, hpars, back, binned=False)

Evaluate Gauss-Hermite 2D function

Form: exp(-a(x-x0)**2 - b(x-x0)*(y-y0) - c(y-y0)**2) + const

Parameters

• X (arraylike [2,npts]) – x,y positions to evaluate at

• x0 (float) – coefficients of function

• y0 (float) – coefficients of function

• a (float) – coefficients of function

• b (float) – coefficients of function

• c (float) – coefficients of function

• back (float) – coefficients of function

pyvista.image.abc2fwxfwytheta(a, b, c)

Convert a,b,c, from gauss2d to xfwhm, yfwhm, theta

pyvista.image.fwxfwytheta2abc(xfwhm, yfwhm, theta)

Convert xfwhm, yfwhm, theta to a,b,c for gauss2d

pyvista.image.gfit2d(data, x0, y0, size=5, fwhm=3.0, sub=True, plot=None, fig=1, scale=1, pafixed=False, astropy=True, binned=False)

Does gaussian fit to input data given initial xcen,ycen

pyvista.image.ghfit2d_thread(pars)

Wrapper for ghfit2d for multithreading, with just a single argument

pyvista.image.ghfit2d(data, x0, y0, size=5, fwhm=3.0, nherm=6, sub=True, plot=None, fig=1, scale=1, pafixed=False, binned=False, p0=None, bounds=None)

Does 2D Gauss-Hermite fit to input data given initial xcen,ycen

pyvista.image.fit2d(X, Y, Z)

Fit 2D quadratic surface

pyvista.image.mk2d(X, Y, coeff)

Return 2D surface given input points and coefficients

pyvista.image.window(data, box, header=None)

Reduce size of image and header accordingly

pyvista.image.stretch(a, ncol=None, nrow=None)

Stretches a 1D image into a 2D image along rows or columns

pyvista.image.add(a, b, dc=0, dr=0, box=None)

Adds b to a, paying attention to CNPIX

pyvista.image.sub(a, b, dc=0, dr=0, box=None)

Subracts b from a, paying attention to CNPIX

pyvista.image.mul(a, b, dc=0, dr=0, box=None)

Multiplies b by a, paying attention to CNPIX

pyvista.image.div(a, b, dc=0, dr=0, box=None)

Divides a by b, paying attention to CNPIX

pyvista.image.clip(hd, min=None, max=None, vmin=None, vmax=None, box=None)

Clipping tasks: sets all values above or below input values to specified values

Parameters

hd – input HDU

Keyword Arguments

• min= (float) – clip values below min

• vmin= (float) – values to clip min values to. If min= is not given clips values <vmin to vmin

• max= (float) – clip values above max

• vmax= (float) – values to clip max values to. If max= is not given clips values >vmax to vmax

pyvista.image.buf(hd)

Display information about HDU

pyvista.image.rd(file, ext=0)

Read file into HDU

pyvista.image.create(box=None, n=None, nr=None, nc=None, sr=1, sc=1, cr=None, cc=None, const=None)

Creates a new HDU

pyvista.image.sky(im, box=None, max=None, min=None, plot=None)

Estimate sky value in an image by fitting parabola to peak of histogram

Parameters

im (HDU or numpy array) – input image data

Keyword Arguments

• box= – only use values within specified box (default=None)

• min= – ignore values below min in sky computation (default=None)

• max= – ignore values above max in sky computation (default=None)

• plot= – matplotlib axes to view histogram and fit (default=None)

pyvista.image.xcorr(a, b, lags, medfilt=0, rad=3)

Cross correlation function between two arrays, calculated at lags

If input images have the same number of rows, then calculate a single

cross-correlation in columns

If first image has one row, but the second has more, then calculate

a cross correlation for each row of the second images

Parameters

• a – array_like reference array

• b – array_like array to calculate shifts for

• lags – array_like x-corrlation lags to use

• medfilt – int, default=0 size of median filter for arrays

Returns :

fit peak of cross-correlation (quadratic fit) 1D cross-correlation function

pyvista.image.xcorr2d(a, b, lags=None, xlags=None, ylags=None)

Two-dimensional cross correlation

Parameters

• a – input Data frames

• b – input Data frames

• lags – array (1D) of x-corrlation lags

Returns

(x,y) position of cross correlation peak from quadratic fit to x-correlation 2D cross correlation function

pyvista.image.zap(hd, size, nsig=3, mask=False)

Median filter array and replace values > nsig*uncertainty

pyvista.image.smooth(data, size, uncertainty=None, bitmask=None)

Boxcar smooth image

pyvista.image.minmax(data, mask=None, low=3, high=10)

Return min,max scaling factors for input data using median, and MAD

Parameters

• img – input CCDData

• low – number of MADs below median to return

• high – number of MADs above median to retunr

Returns

low and high scaling factors

Return type

min,max

pyvista.image.transform(im0, im, lines0, xlags=range(-11, 12), ylags=range(-17, 18), rad=2, scale=20, hard=None, reject=1)

Get geometric transformation between images based on point sources

Parameters

• im0 (Data or array-like) – Reference image

• im (Data or array-like) – Target image

• lines0 (Table) – Table with reference object positions (‘x’ and ‘y’)

• xlags (range) – Range of cross-correlation shifts to try, default is range(-11,12) and (-17,18)

• ylags (range) – Range of cross-correlation shifts to try, default is range(-11,12) and (-17,18)

• rad (float, default=2) – radius for automark

• scale (integer, default=20) – scale for quiver plots

• hard (None or char) – if char, make tranformation plots (’’ to display, otherwise save to specified file name)

pyvista.image.seq(files, box=None, size=None, red=None)

Create montage image from a sequence

Parameters

• files (array-like) – list of files to montage

• box (box, default=None) – image.BOX to use to extract subregion

• size (int, default=None) – if specified, extract region of given size around maximum pixel

• red (Reducer, default=None) – imred.Reducer used to extract images, else use iio.imread