Process Module

This module contains the utilities to parse descriptor file, load image data, process image data to generate results according to the EMVA1288 standart. Results can then be plotted according to the standart.

data

Transform image data into data This class takes a dictionary with image data (product of emva1288.process.loader.LoadImageData) and transforms it into data that can be used for the EMVA1288 computations. It is important to note, that this is separate from LoadImageData because this step, produces float values that are not easily transportable (db, json, etc…) without loosing accuracy.

class emva1288.process.data.Data1288(data, pixels=None, loglevel=20)[source]

Bases: object

Take data from parsed images (descriptor file) and expose it as temporal and spatial dicts This dicts have the appropriate form for the processing

__init__(data, pixels=None, loglevel=20)[source]

Data processing object init method.

It sets the logging configuration and creates a python:logging.Logger object using python:logging.getLogger() and sets the loglevel. Creates the data dictionaries with the call of _get_temporal() for temporal data and _get_spatial() for spatial data.

Parameters:

  • data (The data descriptor file.) –

  • loglevel (int, optional) – The loglevel for the Logger object.

_get_spatial(data)[source]

Fill the spatial dictionary.

The images (sum and pvar) are preserved, they are needed for processing.

Parameters:

data (The data dictionary to take data from.) –

Returns:

dict – The keys are the following:

  • ’texp’: the array of exposure times for spatial tests,

  • ’u_p’: the array of photon count average for each exposure times.

  • ’sum’: the array of the image sum for each photon count and exposure time,

  • ’pvar’: the array of the pvar image for each photon count and exposure time,

  • ’L’: the number of image taken to make the sum and pvar images for each photon count and each exposure time,

  • ’avg’: the average computed from the sum image for each photon count and each exposure time,

  • ’var’: the variance computed from the pvar image for each photon count and each exposure time,

  • ’sum_dark’: the sum image in the dark for each exposure time,

  • ’pvar_dark’: the pvar image in the dark for each exposure time,

  • ’L_dark’: the number of image taken in the dark to compose the sum and pvar images for each exposure time,

  • ’avg_dark’: the average computed from the dark sum image for each exposure time,

  • ’var_dark’: the variance computed from the dark pvar image for each exposure time.

Return type:

A dict containing all spatial test data.

Raises:

ValueError – If the there is no exactly one dark and one bright point

_get_spatial_data(d, postfix='')[source]

Add the mean and variance to the spatial image data.

The mean is the sum of the images divided by L, which is the number of images for the spatial test.

The variance is the pseudovariance divided by (L^2 * (L-1)).

Parameters:

  • d (dict) – The data dictionary containing the sum and pvar of the images.

  • postfix (str, optional) – String to add in the resulting dictionary keys

Returns:

  • dict (A data dictionary processed from the input. The keys (+ postfix))

  • are –

    • ‘sum’: the sum image preserved from input,

    • ’L’: the number of image summed,

    • ’M’: the number of rows for one image

    • ’N’: the number of columns for one image

    • ’avg_var’: variance of the the average image

    • ’avg_mean’: mean of the the average image

    • ’var_mean’: the variance of the variance image

    • ’avg_var_cav’: residual column spatial variance

    • ’avg_var_rav’: residual row spatial variance

_get_temporal(data)[source]

Fill the temporal dict, with the stuff that we need. Compute the averages and variances from the sums (sum and pvar)

If there is only one exposure time, the arrays in the returned dict will all have the same length as the photon count array. For this case, the exposure times and the dark value data array elements will all be the same.

Parameters:

data (The data dictionary containing the temporal data sets.) –

Returns:

dict – The keys are the following: ‘texp’: the array of the exposure times used for the test, ‘u_p’: the array of photon count in a pixel for each exposure time and photon count, ‘u_y’: the array of the mean digital value for each exposure time and photon count, ‘s2_y’: the array of the digital value variance for each exposure time and photon count, ‘u_ydark’: the array of the mean digital dark value for each exposure time and ‘s2_ydark’: the array of the digital dark value variance for each exposure time.

Return type:

A dict containing all temporal test data.

Raises:

ValueError – If there is no one 0.0 photon count entry for each exposure time and at least one bright point

_get_temporal_data(d)[source]

Convert temporal image data to mean and variance

The mean is the sum of the pixels of the two images divided by (2 * self.pixels)

The variance is the pseudo variance(integer), divided by (4 * self.pixels)

The difference of mean values is calculated by difference of two images, used for correction for the variance, squared and then divided by 2

Parameters:

d (dict) – The data dictionary that contains the sum and pvar of the pixels of a sum of two consecutive images with the same photon count and exposure time.

Returns:

dict – ‘mean’: the mean as described above and ‘var’: the variance as described above. ‘dmean’: the mean difference as described above.

Return type:

A data dictionary with the following keys:

loader

Load image data This class takes a dictionary (product of parser.ParseEmvaDescriptorFile). Load the related images and reduce it’s data to the minimum possible, preserving all relevant image data in as integers

class emva1288.process.loader.LoadImageData(images, path='', loglevel=20, fload=<function load_image>, fload_args=[], fload_kwargs={})[source]

Bases: object

Using an info dictionary, load the images and fill data with the relevant information.

__init__(images, path='', loglevel=20, fload=<function load_image>, fload_args=[], fload_kwargs={})[source]

Loader init method.

Uses a python:logging.Logger object to print infos for the user. In the initialization, it loads the infos from the images into the data dictionary.

Parameters:

  • images (dict) – The images attribute of a ParseEmvaDescriptorFile object that have parsed the images of a descriptor file.

  • path (str, optional) – The path to the directory containing the images.

  • loglevel (int, optional) – The information level for the Logger.

  • fload (func. optional) – The function that will load the images (one at a time).

  • fload_args (list, optional) – The list of args for the fload function.

  • fload_kwargs (dict, optional) – The kwargs dictionary for the fload function.

parser

EMVA 1288 descriptor parser This class takes an EMVA1288 descriptor file and loads its content into a python dictionary.

An EMVA1288 descriptor file is a file that contains the description of an EMVA1288 test including exposure times, photon count and corresponding images

class emva1288.process.parser.ParseEmvaDescriptorFile(filename, path=None, loglevel=20)[source]

Bases: object

Take an image descriptor file and transform it into an usable directory

__init__(filename, path=None, loglevel=20)[source]

Parser init method.

Uses a python:logging.Logger object to print infos of the parcing process. This method loads the file and gets the information from it.

Parameters:

  • filename (str) – The descriptor file’s name or the complete path to it.

  • path (str, optional) – The path to the descriptor file.

  • loglevel (int, optional) – The logger level.

plotting

Plot the results This class takes a results.Results1288 object and produces all the plots needed to create a reference datasheet of the EMVA1288 test

emva1288.process.plotting.EVMA1288plots = [<class 'emva1288.process.plotting.PlotPTC'>, <class 'emva1288.process.plotting.PlotSNR'>, <class 'emva1288.process.plotting.PlotSensitivity'>, <class 'emva1288.process.plotting.PlotUyDark'>, <class 'emva1288.process.plotting.PlotLinearity'>, <class 'emva1288.process.plotting.PlotDeviationLinearity'>, <class 'emva1288.process.plotting.PlotHorizontalSpectrogramPRNU'>, <class 'emva1288.process.plotting.PlotHorizontalSpectrogramDSNU'>, <class 'emva1288.process.plotting.PlotVerticalSpectrogramPRNU'>, <class 'emva1288.process.plotting.PlotVerticalSpectrogramDSNU'>, <class 'emva1288.process.plotting.PlotLogarithmicHistogramDSNU'>, <class 'emva1288.process.plotting.PlotLogarithmicHistogramPRNU'>, <class 'emva1288.process.plotting.PlotAccumulatedLogHistogramDSNU'>, <class 'emva1288.process.plotting.PlotAccumulatedLogHistogramPRNU'>, <class 'emva1288.process.plotting.PlotHorizontalProfile'>, <class 'emva1288.process.plotting.PlotVerticalProfile'>]

This list is quite exhaustive. There are the links to corresponding documentation:

class emva1288.process.plotting.Emva1288Plot(figure)[source]

Bases: object

Base class for emva plots.

name = ''

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

rearrange()[source]

Opportunity to change axis or limits after all the tests have been plotted.

Uses matplotlib:matplotlib.figure.Figure.tight_layout() method.

reduce_ticks(ax, axis, n=4)[source]

Reduce the number of ticks in ax.axis

Uses the matplotlib:matplotlib.axes.Axes.locator_params() method.

Parameters:

  • ax (The matplotlib:matplotlib.axes.Axes object to which) – the number of ticks will be changed.

  • axis (str, {'x', 'y', 'both'}) – Axis on which to operate.

  • n (int, optional) – Number of bins between ticks to be left.

set_legend(ax)[source]

Shortcut to add legend.

Parameters:

ax (The matplotlib:matplotlib.axes.Axes object to which) – the legend will be added.

setup_figure()[source]

Simple wrapper for one plot per figure

Takes the name, xlabel, ylabel, xscale and yscale for one plot case.

If more than one plot, just overwrite as you wish.

title = None

The figure’s title.

xlabel = None

The x axis label.

xscale = None

The x axis scale.

ylabel = None

The y axis label.

yscale = None

The y axis scale.

class emva1288.process.plotting.PlotAccumulatedLogHistogramDSNU(figure)[source]

Bases: Emva1288Plot

Create Accumulated log histogram DSNU plot

name = 'Accumulated log histogram DSNU'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

xlabel = 'Minimal deviation from the mean [DN]'

The x axis label.

ylabel = 'Number of pixels'

The y axis label.

yscale = 'log'

The y axis scale.

class emva1288.process.plotting.PlotAccumulatedLogHistogramPRNU(figure)[source]

Bases: Emva1288Plot

Create Accumulated log histogram PRNU plot

name = 'Accumulated log histogram PRNU'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

xlabel = 'Minimal deviation from the mean [%]'

The x axis label.

ylabel = 'Number of pixels'

The y axis label.

yscale = 'log'

The y axis scale.

class emva1288.process.plotting.PlotDeviationLinearity(figure)[source]

Bases: Emva1288Plot

Create Deviation Linearity plot

name = 'Deviation linearity'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

xlabel = '$\\mu_{p}$ [mean number of photons/pixel]'

The x axis label.

ylabel = 'Linearity error LE [%]'

The y axis label.

class emva1288.process.plotting.PlotHorizontalProfile(*args, **kwargs)[source]

Bases: ProfileBase

Create Horizontal profile plot Profile is done with spatial images.

name = 'Horizontal profile'

The figure’s name (used as title if title is none).

rearrange()[source]

Opportunity to change axis or limits after all the tests have been plotted.

Uses matplotlib:matplotlib.figure.Figure.tight_layout() method.

setup_figure()[source]

Simple wrapper for one plot per figure

Takes the name, xlabel, ylabel, xscale and yscale for one plot case.

If more than one plot, just overwrite as you wish.

vertical = False
class emva1288.process.plotting.PlotHorizontalSpectrogramDSNU(figure)[source]

Bases: Emva1288Plot

Create Horizontal spectrogram DSNU plot

name = 'Horizontal spectrogram DSNU'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

xlabel = 'cycles [periods/pixel]'

The x axis label.

ylabel = 'Standard deviation and\nrelative presence of each cycle [DN]'

The y axis label.

yscale = 'log'

The y axis scale.

class emva1288.process.plotting.PlotHorizontalSpectrogramPRNU(figure)[source]

Bases: Emva1288Plot

Create Horizontal spectrogram PRNU plot

name = 'Horizontal spectrogram PRNU'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

xlabel = 'cycles [periods/pixel]'

The x axis label.

ylabel = 'Standard deviation and\nrelative presence of each cycle [%]'

The y axis label.

yscale = 'log'

The y axis scale.

class emva1288.process.plotting.PlotLinearity(figure)[source]

Bases: Emva1288Plot

Create Linearity plot

name = 'Linearity'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

xlabel = '$\\mu_{p}$ [mean number of photons/pixel]'

The x axis label.

ylabel = '$\\mu_y - \\mu_{y.dark}$ [DN]'

The y axis label.

class emva1288.process.plotting.PlotLogarithmicHistogramDSNU(figure)[source]

Bases: Emva1288Plot

Create Logarithmic histogram DSNU plot

name = 'Logarithmic histogram DSNU'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

xlabel = 'Dark value [DN]'

The x axis label.

ylabel = 'Number of pixels'

The y axis label.

yscale = 'log'

The y axis scale.

class emva1288.process.plotting.PlotLogarithmicHistogramPRNU(figure)[source]

Bases: Emva1288Plot

Create Logarithmic histogram PRNU plot

name = 'Logarithmic histogram PRNU'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

xlabel = 'Deviation from the mean [DN]'

The x axis label.

ylabel = 'Number of pixels'

The y axis label.

yscale = 'log'

The y axis scale.

class emva1288.process.plotting.PlotPTC(figure)[source]

Bases: Emva1288Plot

Create Photon Transfer plot

name = 'Photon Transfer'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

xlabel = '$\\mu_y - \\mu_{y.dark}$ [DN]'

The x axis label.

ylabel = '$\\sigma^2_y - \\sigma^2_{y.dark}$ [DN$^2$]'

The y axis label.

class emva1288.process.plotting.PlotSNR(figure)[source]

Bases: Emva1288Plot

Create SNR plot

name = 'Signal to Noise Ratio'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

setup_figure()[source]

Simple wrapper for one plot per figure

Takes the name, xlabel, ylabel, xscale and yscale for one plot case.

If more than one plot, just overwrite as you wish.

xlabel = '$\\mu_{p}$ [mean number of photons/pixel]'

The x axis label.

ylabel = 'SNR'

The y axis label.

class emva1288.process.plotting.PlotSensitivity(figure)[source]

Bases: Emva1288Plot

name = 'Sensitivity'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

rearrange()[source]

Opportunity to change axis or limits after all the tests have been plotted.

Uses matplotlib:matplotlib.figure.Figure.tight_layout() method.

xlabel = '$\\mu_p$ [mean number of photons/pixel]'

The x axis label.

ylabel = '$\\mu_y - \\mu_{y.dark}$ [DN]'

The y axis label.

class emva1288.process.plotting.PlotUyDark(figure)[source]

Bases: Emva1288Plot

Create $mu_{y.dark}$ plot

name = 'Mean gray value in dark'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

title = '$\\mu_{y.dark}$'

The figure’s title.

xlabel = 'exposure time [ns]'

The x axis label.

ylabel = '$\\mu_{y.dark}$ [DN]'

The y axis label.

class emva1288.process.plotting.PlotVerticalProfile(*args, **kwargs)[source]

Bases: ProfileBase

Create Vertical profile plot. Profile is done with spatial images.

name = 'Vertical profile'

The figure’s name (used as title if title is none).

rearrange()[source]

Opportunity to change axis or limits after all the tests have been plotted.

Uses matplotlib:matplotlib.figure.Figure.tight_layout() method.

setup_figure()[source]

Simple wrapper for one plot per figure

Takes the name, xlabel, ylabel, xscale and yscale for one plot case.

If more than one plot, just overwrite as you wish.

vertical = True
class emva1288.process.plotting.PlotVerticalSpectrogramDSNU(figure)[source]

Bases: Emva1288Plot

Create Vertical spectrogram DSNU plot

name = 'Vertical spectrogram DSNU'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

xlabel = 'Cycles [periods/pixel]'

The x axis label.

ylabel = 'Standard deviation and\nrelative presence of each cycle [DN]'

The y axis label.

yscale = 'log'

The y axis scale.

class emva1288.process.plotting.PlotVerticalSpectrogramPRNU(figure)[source]

Bases: Emva1288Plot

Create Vertical spectrogram PRNU plot

name = 'Vertical spectrogram PRNU'

The figure’s name (used as title if title is none).

plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

xlabel = 'cycles [periods/pixel]'

The x axis label.

ylabel = 'Standard deviation and\nrelative presence of each cycle [%]'

The y axis label.

yscale = 'log'

The y axis scale.

class emva1288.process.plotting.Plotting1288(*tests)[source]

Bases: object

EMVA1288 plots

Creates and shows all plots necessary to prepare a camera or sensor descriptive report compliant with EMVA Standard 1288.

plot(*plots)[source]

Plot EMVA1288 plots

Parameters:

plots (list) – List of plots to plot

class emva1288.process.plotting.ProfileBase(*args, **kwargs)[source]

Bases: Emva1288Plot

get_profiles(bright, dark)[source]
plot(test)[source]

Method to show the figures.

Parameters:

test (Do nothing for this method but can be) – used for the subclass method.

Raises:

NotImplementedError – If this method is not overridden.

Notes

Must be overridden in subclasses.

results

Compute EMVA1288 values from data This class takes the data from data.Data1288 and compute the actual EMVA1288 values.

class emva1288.process.results.Results1288(data, pixel_area=None, index_u_ysat=None, loglevel=10)[source]

Bases: object

Class used to process data and to generate pdf report using LaTeX.

When properties from this class are computed, their docstring are also parsed using the cls_1288_info() function to retrieve more data informations (like units or full name).

property DR

Dynamic Range.

Defined as the saturation capacity divided by the absolute sensitivity threshold. The greater this number is, the greater the operational range of a camera (between the dark noise level and the saturation level).

Emva1288

  • section: sensitivity
  • short: Dynamic Range
  • symbol: DR
  • latexname: DR
DR_dB()[source]

Dynamic Range in deciBels.

It is defined as 20 * log_10 ( Dynamic Range ).

Emva1288

  • section: sensitivity
  • short: Dynamic Range
  • symbol: $DR_{dB}$
  • unit: dB
  • latexname: DRDB
property DSNU1288

DSNU overall.

Dark Signal NonUniformity (in e^-) is defined as the standard deviation of the dark signal divided by the overall system gain. If the variance is negative, it will return NaN instead of an imaginary number.

Emva1288

  • section: spatial
  • short: DSNU
  • symbol: $DSNU_{1288}$
  • unit: $e^-$
  • latexname: DSNU
DSNU1288_DN()[source]

DSNU in DN.

Defined as the DSNU in e^- multiplied by the overall system gain. Returns NaN if the dark signal variance is negative.

Returns:

float

Return type:

The DSNU in DN.

Emva1288

  • section: spatial
  • short: DSNU in DN
  • symbol: $DSNU_{1288.DN}$
  • unit: DN
  • latexname: DSNUDN
property DSNU1288_col

DSNU in columns.

Dark Signal NonUniformity (in e^-) is defined as the standard deviation of the dark signal divided by the overall system gain. If the variance is negative, it will return NaN instead of an imaginary number.

Emva1288

  • section: spatial
  • short: DSNUcol
  • symbol: $DSNU_{1288.col}$
  • unit: $e^-$
  • latexname: DSNUcol
property DSNU1288_pixel

DSNU in pixel.

Dark Signal NonUniformity (in e^-) is defined as the standard deviation of the dark signal divided by the overall system gain. If the variance is negative, it will return NaN instead of an imaginary number.

Emva1288

  • section: spatial
  • short: DSNUpixel
  • symbol: $DSNU_{1288.pixel}$
  • unit: $e^-$
  • latexname: DSNUpixel
property DSNU1288_row

DSNU in rows.

Dark Signal NonUniformity (in e^-) is defined as the standard deviation of the dark signal divided by the overall system gain. If the variance is negative, it will return NaN instead of an imaginary number.

Emva1288

  • section: spatial
  • short: DSNUrow
  • symbol: $DSNU_{1288.row}$
  • unit: $e^-$
  • latexname: DSNUrow
property K

Overall system gain.

Slope of (s2_y - s2_y_dark) Vs (u_y - u_y_dark). Fit with offset = 0. Uses the LinearB0() to make the fit.

Emva1288

  • section: sensitivity
  • short: System gain
  • symbol: K
  • unit: $DN/e^-$
  • latexname: K
property LE_max

Max Linearity error.

Emva1288

  • section: linearity
  • short: Max Linearity error
  • symbol: $LE_{max}$
  • unit: \%
  • latexname: LEMax
property LE_min

Min Linearity error.

Emva1288

  • section: linearity
  • short: Min Linearity error
  • symbol: $LE_{min}$
  • unit: \%
  • latexname: LEMin
property PRNU1288

PRNU overall.

Photo Response NonUniformity (in %) is defined as the square root of the difference between the spatial variance of a bright image (or from an average of bright images to remove temporal difformities) and the spatial variance of dark signal, divided by the difference between the mean of a bright image and the mean of a dark image.

Emva1288

  • section: spatial
  • short: PRNU
  • symbol: $PRNU_{1288}$
  • unit: \%
  • latexname: PRNU
property PRNU1288_col

PRNU in column.

Photo Response NonUniformity (in %) is defined as the square root of the difference between the spatial variance of a bright image (or from an average of bright images to remove temporal difformities) and the spatial variance of dark signal, divided by the difference between the mean of a bright image and the mean of a dark image.

Emva1288

  • section: spatial
  • short: PRNUcol
  • symbol: $PRNU_{1288.col}$
  • unit: \%
  • latexname: PRNUcol
property PRNU1288_pixel

PRNU in pixel.

Photo Response NonUniformity (in %) is defined as the square root of the difference between the spatial variance of a bright image (or from an average of bright images to remove temporal difformities) and the spatial variance of dark signal, divided by the difference between the mean of a bright image and the mean of a dark image.

Emva1288

  • section: spatial
  • short: PRNUpixel
  • symbol: $PRNU_{1288.pixel}$
  • unit: \%
  • latexname: PRNUpixel
property PRNU1288_row

PRNU in row.

Photo Response NonUniformity (in %) is defined as the square root of the difference between the spatial variance of a bright image (or from an average of bright images to remove temporal difformities) and the spatial variance of dark signal, divided by the difference between the mean of a bright image and the mean of a dark image.

Emva1288

  • section: spatial
  • short: PRNUrow
  • symbol: $PRNU_{1288.row}$
  • unit: \%
  • latexname: PRNUrow
property QE

Quantum efficiency.

It is retrieved as the ratio of the responsivity to the overall system gain.

Emva1288

  • section: sensitivity
  • short: Quantum efficiency
  • symbol: $\eta$
  • unit: \%
  • latexname: QE
property R

Responsivity.

Slope of the (u_y - u_y_dark) Vs u_p. Fit with offset = 0 Uses the LinearB0() function to make the fit.

Emva1288

  • section: sensitivity
  • short: Responsivity
  • symbol: R
  • unit: DN/p
property SNR_max

Maximum Signal-to-Noise Ratio.

Emva1288

  • section: sensitivity
  • short: Signal-to-Noise Ratio
  • symbol: $SNR_{max}$
  • latexname: SNRMax
SNR_max_bit()[source]

Maximum Signal to Noise Ratio in Bits.

Emva1288

  • section: sensitivity
  • short: Maximum Signal to Noise Ratio in Bits
  • symbol: $SNR_{max.bit}$
  • unit: bit
  • latexname: SNRMaxBit
SNR_max_dB()[source]

Maximum Signal to Noise Ratio in Db.

Emva1288

  • section: sensitivity
  • short: Maximum Signal to Noise Ratio in Db
  • symbol: $SNR_{max.dB}$
  • unit: dB
  • latexname: SNRMaxDB
__init__(data, pixel_area=None, index_u_ysat=None, loglevel=10)[source]

Results computation init method.

This class uses a python:logging.Logger object to display informations for users.

Parameters:

  • data (dict) – The data dictionary to compute the results from.

  • pixel_area (float, optional) – The area of one pixel in um^2.

  • index_u_ysat (int, optional) – The index of the u_y array at which we consider that the camera saturates. This is used if forcing the saturation point is necessary.

  • loglevel (int, optional) – The level for the python:logging.Logger object.

property histogram_DSNU

DSNU histogram.

Uses the Histogram1288() function to make the histogram.

Emva1288

  • section: defect_pixel
  • short: DSNU histogram
property histogram_DSNU_accumulated

Accumulated DSNU histogram.

Uses the Histogram1288() function to make the histogram.

Emva1288

  • section: defect_pixel
  • short: accumulated DSNU histogram
property histogram_PRNU

PRNU histogram.

Uses the high_pass_filter() function to filter the image and then uses the Histogram1288() function to make the histogram.

Emva1288

  • section: defect_pixel
  • short: PRNU histogram
property histogram_PRNU_accumulated

Accumulated PRNU histogram.

Uses the high_pass_filter() function to filter the image and then uses the Histogram1288() function to make the histogram.

Emva1288

  • section: defect_pixel
  • short: accumulated PRNU histogram
property index_linearity_max

Linearity fit maximum index.

Maximum index for linear fit (95% of saturation).

Emva1288

  • section: linearity
  • short: Linearity fit maximum index
property index_linearity_min

Linearity fit minimun index.

Minimum index for linear fit (5% of saturation).

Emva1288

  • section: linearity
  • short: Linearity fit minimun index
property index_sensitivity_max

Index for linear fits in sensitivity part of the standard (70% of saturation).

Emva1288

  • section: sensitivity
  • short: Sensitivity fit maximum index
property index_sensitivity_min

Sensitivity minimum index.

Index for linear fits in sensitivity part of the standard (70% of saturation)

Emva1288

  • section: sensitivity
  • short: Sensitivity fit minimum index
property index_start

The array’s starting index.

Emva1288

  • section: info
  • short: Start array index
property index_u_ysat

Index of saturation.

Emva1288

  • section: sensitivity
  • short: Saturation index
inverse_K()[source]

Inverse of overall system gain.

Emva1288

  • section: sensitivity
  • short: Inverse of overall system gain
  • symbol: 1/K
  • unit: $e^-/DN$
  • latexname: InvK
inverse_SNR_max()[source]

Inverse Maximum Signal to Noise Ratio.

Emva1288

  • section: sensitivity
  • short: Inverse Maximum Signal to Noise Ratio
  • symbol: $SNR_{max}^{-1}$
  • unit: \%
  • latexname: InvSNRMax
linearity()[source]

Returns a dictionary containing linearity information.

It fits the mean digital signal in function of the mean photon count (Linear fit) using the EMVA1288 standard for linear fit.

Returns:

dict – The keys are:

  • ’fit_slope’ : The slope of the linear fit.

  • ’fit_offset’ : The offset of the fit.

  • ’relative_deviation’ : The relative deviation of the real data from the fit (in %) for the whole array.

  • ’linearity_error_min’ : The minimal value of the relative deviation.

  • ’linearity_error_max’ : The maximal value of the relative deviation.

Return type:

Linearity dictionary.

print_results()[source]

Print results to the screen.

property results

Dictionnary with all the values and metadata for EMVA1288 values.

It uses the obj_to_dict() to compute all the results at once.

property results_by_section

Results ordered by section.

property s2q

Quantification noise.

property s_2_y

Spatial variance from image.

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y}$
  • unit: DN2
property s_2_y_cav

Average of column spatial variance from image.

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y.cav}$
  • unit: DN2
property s_2_y_cav_dark

Average of column spatial variance from image.

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y.cav.dark}$
  • unit: DN2
property s_2_y_col

Column spatial variance from image.

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y.col}$
  • unit: DN2
property s_2_y_col_dark

Column spatial variance from dark image.

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y.col.dark}$
  • unit: DN2
property s_2_y_dark

Spatial variance from image,

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y}$
  • unit: DN2
property s_2_y_measured

Spatial variance measure.

Variance value of the bright variance image.

Emva1288

  • section: spatial
  • short: Spatial variance measure
  • symbol: $s^2_{y.measured}$
  • unit: DN2
property s_2_y_measured_dark

Spatial variance measured dark.

Variance value of the dark variance image.

Emva1288

  • section: spatial
  • short: Spatial variance measured dark
  • symbol: $s^2_{y.measured.dark}$
  • unit: DN2
property s_2_y_pixel

Pixel spatial variance from image.

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y.pixel}$
  • unit: DN2
property s_2_y_pixel_dark

Pixel spatial variance from dark image.

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y.pixel.dark}$
  • unit: DN2
property s_2_y_rav

Average of row spatial variance from image.

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y.rav}$
  • unit: DN2
property s_2_y_rav_dark

Average of row spatial variance from image.

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y.rav.dark}$
  • unit: DN2
property s_2_y_row

Row spatial variance from image.

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y.row}$
  • unit: DN2
property s_2_y_row_dark

Row spatial variance from dark image.

Emva1288

  • section: spatial
  • short: Spatial variance from image
  • symbol: $s^2_{y.row.dark}$
  • unit: DN2
property sigma_2_y_stack

Temporal variance stack.

Mean value of the bright variance image.

Emva1288

  • section: spatial
  • short: Temporal variance stack
  • symbol: $\sigma^2_{y.stack}$
  • unit: DN2
property sigma_2_y_stack_dark

Temporal variance stack dark.

Mean value of the dark variance image.

Emva1288

  • section: spatial
  • short: Temporal variance stack dark
  • symbol: $\sigma^2_{y.stack.dark}$
  • unit: DN2
property sigma_d

Temporal Dark Noise.

Emva1288

  • section: sensitivity
  • short: Temporal Dark Noise
  • symbol: $\sigma_d$
  • unit: $e^-$
  • latexname: SigmaDark
property sigma_y_dark

Temporal Dark Noise.

Uses LinearB() to make the fit.

Emva1288

  • section: sensitivity
  • short: Temporal Dark Noise
  • symbol: $\sigma_{y.dark}$
  • unit: DN
  • latexname: SigmaYDark
property u_I_mean

Dark Current from mean.

The dark current from mean is the slope of the dark signal mean in function of the exposure times divided by the overall system gain. Returns NaN if the number of different exposure times is less than 3.

Emva1288

  • section: dark_current
  • short: Dark Current from mean
  • symbol: $\mu_{I.mean}$
  • unit: $e^-/s$
property u_I_mean_DN

Dark Current from mean.

The dark current from mean is the slope of the dark signal mean in function of the exposure time. Returns NaN if the number of different exposure times is less than 3.

Emva1288

  • section: dark_current
  • short: Dark Current from mean
  • symbol: $\mu_{I.mean.DN}$
  • unit: $DN/s$
property u_I_var

Dark Current from variance.

The dark current from variance (in e-/s) is the dark current from variance (in DN/s) divided by the system gain. Returns NaN if the fit slope is negative.

Emva1288

  • section: dark_current
  • short: Dark Current from variance
  • symbol: $\mu_{I.var}$
  • unit: $e^-/s$
  • latexname: UIVar
property u_I_var_DN

Dark Current from variance.

The dark current from variance (in DN/s) is the slope of the dark signal variance as a function of the exposure time divided by the system gain. Returns NaN if the slope is negative.

Emva1288

  • section: dark_current
  • short: Dark Current from variance
  • symbol: $\mu_{I.var.DN}$
  • unit: $DN/s$
  • latexname: UIVar
property u_e_min

Sensitivity threshold.

Emva1288

  • section: sensitivity
  • short: Sensitivity threshold
  • symbol: $\mu_{e.min}$
  • unit: $e^-$
  • latexname: UEMin
property u_e_min_area

Sensitivity threshold per pixel area.

Returns None if the pixel area is not defined or 0.

Emva1288

  • section: sensitivity
  • short: Sensitivity threshold
  • symbol: $\mu_{e.min.area}$
  • unit: $e^-/\mu m^2$
  • latexname: UEMin
property u_e_sat

Saturation Capacity.

Number of electrons at saturation.

Emva1288

  • section: sensitivity
  • short: Saturation Capacity
  • symbol: $\mu_{e.sat}$
  • unit: $e^-$
  • latexname: UESat
property u_e_sat_area

Saturation Capacity per pixel area.

Returns None if pixel area is not defined or 0.

Emva1288

  • section: sensitivity
  • short: Saturation Capacity
  • symbol: $\mu_{e.sat.area}$
  • unit: $e^-/\mu m^2$
  • latexname: UESat
property u_p_min

Absolute sensitivity threshold.

Emva1288

  • section: sensitivity
  • short: Absolute sensitivity threshold
  • symbol: $\mu_{p.min}$
  • unit: $p$
  • latexname: UPMin
property u_p_min_area

Sensitivity threshold per pixel area.

Returns None if pixel area is not defined or 0.

Emva1288

  • section: sensitivity
  • short: Sensitivity threshold
  • symbol: $\mu_{p.min.area}$
  • unit: $p/\mu m^2$
  • latexname: UPMin
property u_p_sat

Saturation Capacity.

Emva1288

  • section: sensitivity
  • short: Saturation Capacity
  • symbol: $\mu_{p.sat}$
  • unit: $p$
  • latexname: UPSat
property u_p_sat_area

Saturation Capacity per pixel area.

Returns None if pixel area is not defined or 0.

Emva1288

  • section: sensitivity
  • short: Saturation Capacity
  • symbol: $\mu_{p.sat.area}$
  • unit: $p/\mu m^2$
  • latexname: UPSat
xml(filename=None)[source]

Method that writes the results in xml format to a file.

Parameters:

filename (str, optional) – The file to write the results. If None, the xml string won’t be written but will be returned instead.

Returns:

str

Return type:

If the xml string is not written into a file, it is returned.

routines

Utils functions

emva1288.process.routines.FFT1288(img, rotate=False, n=1)[source]

Compute the FFT emva1288 style

Compute an FFT per line and average the resulting ffts

If img is a masked array then remove the masked data line by line and calculate the FFT on the minimum number of valid columns.

Parameters:

  • img (array) – Input image

  • rotate (bool (optional)) – Rotate the image before performing the FFT

  • n (int (optional)) – If the image is the sum of several images use this value to produce the fft of the average image

Returns:

array

Return type:

One dimension FFT power spectrum

emva1288.process.routines.GetFrecs(fft)[source]
emva1288.process.routines.GetImgShape(img)[source]
emva1288.process.routines.Histogram1288(img, Qmax)[source]
emva1288.process.routines.LinearB(Xi, Yi)[source]
emva1288.process.routines.LinearB0(Xi, Yi)[source]
emva1288.process.routines.cls_1288_info(cls)[source]

Dictionnary that represents results.

Parameters:

cls (Class from wich to extract the information.) –

Returns:

Dictionnary extracted using the format defined by a custom sphinx directive with the following format:

{attribute1: {'section': section name,
              'units': attribute units,
              'short': attribute short description,
              'latexname': latex name for the attribute,
              'symbol': symbol to represent the value}}

Return type:

dict

emva1288.process.routines.compare_xml(x1, x2, filename=None)[source]
emva1288.process.routines.dict_to_xml(d, root='results', filename=None)[source]

Takes a dict and return a well formed xml string

emva1288.process.routines.get_int_imgs(imgs)[source]

Returns the sum, pseudo-variance, mean difference from list of images sum is just the image resulting on the addition of all the images pvar is the pseudo-variance, this is defined by: pvar = SUM((L(i)^2 - (SUM(i))^2) to get variance from pseudo-variance var = (1/(L^2) * 1/(L - 1)) * pvar dmean is the mean difference between the images, obtained by: SUM(mean(i) - mean(i-1)), when i>0 this is used to remove mu_y^2 from the total variance important value to get sigma_y (temporal variance).

keys & value in output dict:

L : number of images for computation–> int sum : total summed image [M x N] –> array. int64 pvar : summed variance image [M x N] –> array. int64 dmean: summed mean differences –> float

emva1288.process.routines.high_pass_filter(img, dim)[source]

High pass filtering on image

Computes the highpass filtering as defined by the emva standart keeping the result as an int image. The computation preserves the image as ints and doesn’t perform the final division, but returns the dividing factor

Parameters:

  • img (np.array) – the image to filter

  • dim (int) – The size of the highpass filter

Returns:

d – The data dictionary of the result. The keys are: ‘img’ : The filtered image ‘multiplicator’ : Factor from the convolution to be considered in the computation of the histogram

Return type:

dict

emva1288.process.routines.load_image(fname)[source]
emva1288.process.routines.obj_to_dict(obj)[source]

Get the info dict from the object class Add the values or Data to this dict

for each method if the return value is a dict, it is inserted as d[SectionName][MethodName][Data] = ReturnValue if not d[SectionName][MethodName][Value] = ReturnValue

emva1288.process.routines.round_significant(v, sig=7)[source]

Round up to the given significant digits, used for comparison

emva1288.process.routines.xml_to_dict(xml)[source]

If xml is a file, opens and parse, if string, parse it from string Convert the xml to a dict using element_to_dict Process the resulting dict: Cast Data to numpy float arrays (split the string by whitespaces) Cast Value to float