Презентация на тему EMVA 1288 Standard for measurement and presentation of specifications for machine vision. Sensors and cameras

organization
Measurement setup
1288 Module 1 “in 12 Slides”



Interpretation of the

standard for the user

Agenda

Example DR-4k-7

MV
Each Sensor and Camera Manufacturer has it’s own definitions
The

Customer is left guessing with datasheets
Today cameras and sensor are digital but manufacturers specify V/lux s ?!
Lux is a unit adapted vor human vision not for wider spectral band

L(λ)−curve

QE-Sensor

Illumination

= 0 Lux!!

= ?

EMVA 1288 Working group is open! Contact info@emva.org

Temporal Noise; Linearity
Adapted by many companies in Europe and

World wide for internal use and on customer request.
Published data on www.standard1288.org from
AWAIBA
BASLER
PCO

Supported world wide by agreement between JIIA, AIA & EMVA on G3 standardization frame work

of temperature measurements


Standard requirements
Standard 1288 Module 1 1) Measurement Setup

sphere
Recommended Setup:
(Preliminary)


Diffuser

standard will only produce sense full results for cameras

/ sensors that follow the mathematical model !

Current module covers only linear integrating sensors
More sensor types to follow

general information and information regarding the operation point at

which data is acquired.

Vendor name
Model name
Type of data presented: Typical; Guaranteed; Guaranteed over life time[1]
Sensor type (CCD; CMOS; CID etc...)
Sensor diagonal in [mm]; Indication of lens category to be used [inch]
Resolution; Pixel size (width x height in [µm])
Readout type (CCD only) ; Transfer type (CCD only);
Shutter type (CMOS only); Global; Rolling
Overlap capabilities; (readout of frame n and exposure of frame n+1 can happen at the same time).
Maximum frame rate at the given operation point. (no change of settings permitted)
Others (Interface Type etc..)

exactly in what configuration the camera / Sensor is

operated. I.e. all gain & register settings and environmental conditions.
Unless for the scans of the integration time, the camera / sensor has to be in the same operation point for all EMVA1288 measurements.

of images with increasing Exposure time. One series in

dark, one with a constant illumination. Compute the mean value of The images, report the mean versus the integration time and the number of photons Collected in each pixel.

From 2 images at the same integration time, compute the square difference image. The mean of the squared difference image is the Variance. Report the variance for the series in dark and for the series with light versus integration time and number of photons. Saturation exposure is defined as the exposure where the variance is maximal

mean dark values from the mean values with illumination.
Subtract

the dark variance values form the bright variance values.
Report the dark corrected variance versus the dark corrected mean values.
Fit a line going through the origin through the data points until saturation
The slope of this line gives the conversion gain in DN/e- or K

values resulting from each integration time versus the number

of photons impinging on a pixel during this integration time.
Use Logarithmic scale (dB or bit (base log2)
Plot values from lowest chosen integration time til saturation.

based on
photon shot noise
Of incoming signal
Maximum SNR

conversion gain and the light intensity used during exposure

compute the total QE.
Report the QE versus measurement wavelengths.
Note for EMVA1288 data given per unit: When clearly stated this measurement can be based on typical data even for production data, however the curve should be scaled to the QE measured at the specific measurement wavelength of sample.

dark images versus increasing integration times at different housing

temperatures. (Measurement not applicable for cooled cameras)
Report the increased signal values versus integration time.
Report the increase of variance versus integration time.

Subtract from each series the mean signal value and the variance value of the shortest integration time from each measurements point.
From the increase in variance compute the dark current in [e-/s] and the dark current doubling temperature [K]

current at ambient 3e-/ms

the DFFT
Average the FFT values along each column of

the FFT result matrix
Plot the resulting spectrogram for Dark; 50% Saturation; 90% Saturation

Idea: describe DSNU & PRNU effects in from of a spectral amplitude plot to
Show high frequency (E.g. odd even mismatch) and low frequency
(E.g. over all gradients) parts of spatial and temporal noise separately

visual impression! Raw sample image DR-4k-7 at previously described operation

point

SPECIMEN