Basler ral km. Camera Specification. Measurement protocol using the EMVA Standard 1288 Document Number: BD Version: 01

Transcription

1 Basler ral8-8km Camera Specification Measurement protocol using the EMVA Standard 188 Document Number: BD79 Version: 1

2 For customers in the U.S.A. This equipment has been tested and found to comply with the limits for a Class A digital device, pursuant to Part 15 of the FCC Rules. These limits are designed to provide reasonable protection against harmful interference when the equipment is operated in a commercial environment. This equipment generates, uses, and can radiate radio frequency energy and, if not installed and used in accordance with the instruction manual, may cause harmful interference to radio communications. Operation of this equipment in a residential area is likely to cause harmful interference in which case the user will be required to correct the interference at his own expense. You are cautioned that any changes or modifications not expressly approved in this manual could void your authority to operate this equipment. The shielded interface cable recommended in this manual must be used with this equipment in order to comply with the limits for a computing device pursuant to Subpart J of Part 15 of FCC Rules. For customers in Canada This apparatus complies with the Class A limits for radio noise emissions set out in Radio Interference Regulations. Pour utilisateurs au Canada Cet appareil est conforme aux normes Classe A pour bruits radioélectriques, spécifiées dans le Règlement sur le brouillage radioélectrique. Life Support Applications These products are not designed for use in life support appliances, devices, or systems where malfunction of these products can reasonably be expected to result in personal injury. Basler customers using or selling these products for use in such applications do so at their own risk and agree to fully indemnify Basler for any damages resulting from such improper use or sale. Warranty Note Do not open the housing of the camera. The warranty becomes void if the housing is opened. All material in this publication is subject to change without notice and is copyright Basler AG.

3 Contacting Basler Support Worldwide Europe: Basler AG An der Strusbek Ahrensburg Germany Tel.: Fax.: support.europe@baslerweb.com Americas: Basler, Inc. 855 Springdale Drive, Suite 3 Exton, PA 1931 U.S.A. Tel.: Fax.: support.usa@baslerweb.com Asia: Basler Asia Pte. Ltd. 35 Marsiling Industrial Estate Road 3 # 5-6 Singapore Tel.: Fax.: support.asia@baslerweb.com

4

5 CONTENTS Contents 1 Overview 7 Introduction 8 3 Basic Information Illumination Illumination Setup for the Basler Camera Test Tool Measurement of the Irradiance Characterizing Temporal Noise and Sensitivity 11.1 Basic Parameters Total Quantum Efficiency Temporal Dark Noise Dark Current Doubling Temperature Inverse of Overall System Gain Inverse Photon Transfer Saturation Capacity Derived Data Absolute Sensitivity Threshold Signal-to-noise Ratio Dynamic Range Raw Measurement Data Mean Gray Value Variance of the Temporal Distribution of Gray Values Mean of the Gray Values Dark Signal Variance of the Gray Value Temporal Distribution in Darkness Light Induced Variance of the Temporal Distribution of Gray Values Light Induced Mean Gray Value Dark Current Versus Housing Temperature Characterizing Total and Spatial Noise Basic Parameters Spatial Offset Noise Spatial Gain Noise Raw Measurement Data Standard Deviation of the Spatial Dark Noise Light Induced Standard Deviation of the Spatial Noise Bibliography 33 Basler ral8-8km 5

6 CONTENTS 6 Basler ral8-8km

7 1 Overview 1 Overview Basler ral8-8km Item Symbol Typ. 1 Unit Remarks Temporal Noise Parameters Total Quantum Efficiency (QE) η % λ = 55 nm Inverse of Overall System Gain 1 K 6. e DN Temporal Dark Noise σ d 9 e Saturation Capacity µ e.sat 9 e Derived Parameters Absolute Sensitivity Threshold µ p.min 1 p λ = 55 nm Dynamic Range DYN out.bit 11.5 bit Maximum SNR SNR y.max.bit 7.3 bit Spatial Noise Parameters SNR y.max.db. db Item Symbol Typ. Unit Remarks Spatial Offset Noise, DSNU 188 σ o.3 e Spatial Gain Noise, PRNU 188 S g. % Table 1: Most Important Specification Data Operating Point Item Symbol Remarks Video output format 1 bits/pixel(mono16) Gain Register raw 56 Offset Register raw 3 Exposure time T exp. µs to 3.6 ms Table : Operating Point for the Camera Used 1 The unit e is used in this document as a statistically measured quantity. Basler ral8-8km 7

8 Introduction Introduction This measurement protocol describes the specification of Basler ral8-8km cameras. The measurement methods conform to the 188 EMVA Standard, the Standard for Characterization and Presentation of Specification Data for Image Sensors and Cameras (Release A1.3) of the European Machine Vision Association (EMVA) [1]. The most important specification data for Basler ral8-8km cameras is summarized in table 1. 8 Basler ral8-8km

9 3 Basic Information 3 Basic Information Basic Information Vendor Basler Model ral8-8km Type of data presented Typical Number of samples 1 Sensor Awaiba DR-k-7 Linear Sensor type CMOS Sensor diagonal Indication of lens category to be used F-Mount Resolution 8 pixel Pixel width 7. µm Pixel height 7. µm Readout type Transfer type Shutter type - Overlap capabilities Maximum readout rate 8. khz General conventions - Interface type Camera Link Table 3: Basic Information Basler ral8-8km 9

10 3.1 Illumination 3.1 Illumination Illumination Setup for the Basler Camera Test Tool The illumination during the testing on each camera was fixed. The drift in the illumination over a long period of time and after the lamp is changed is measured by a reference Basler A6fc camera. The reference camera provides an intensity factor that was used to calculate the irradiance for each camera measurement. Light Source Item Symbol Typ. Unit Remarks Wavelength λ 55 nm Wavelength Variation λ 5 nm Distance sensor to light source d 8 mm Diameter of the light source D 35 mm f-number f # 8 f # = d D Table : Light Source 3.1. Measurement of the Irradiance The irradiance was measured using an IL17 Radiometer from International Light Inc. (Detector: SEL33 #685; Input optic: W #961; Filter: F #187; regular calibration). The accuracy of the Radiometer is specified as ±3.5%. The measured irradiance is plotted in figure 1..5 'ral8-8km' (1 cameras), Irradiance Irradiance [W/m^] Measurement Figure 1: Irradiance for Each Camera Measurement. The error for each calculated value using the amount of light falling on the sensor is dependent on the accuracy of the irradiance measurement. 1 Basler ral8-8km

11 Characterizing Temporal Noise and Sensitivity Characterizing Temporal Noise and Sensitivity.1 Basic Parameters.1.1 Total Quantum Efficiency Total Quantum Efficiency for One Fixed Wavelength Total quantum efficiency η(λ) in [%] for monochrome light at λ = 55 nm with a wavelength variation of λ = 5 nm. 5 'ral8-8km' (1 cameras), Quantum Efficiency Quantum Efficiency [%] Camera Number 'ral8-8km' (1 cameras), Quantum Efficiency Histogram 1 3 Quantum Efficiency [%] Figure : Total Quantum Efficiency (QE) Item Symbol Typ. Std. Dev. Unit Remarks Total Quantum Efficiency (QE) η TBD % λ = 55 nm Table 5: Total Quantum Efficiency (QE) The main error in the total quantum efficiency η is related to the error in the measurement of the illumination as described in section 3.1. Basler ral8-8km 11

12 .1 Basic Parameters Total Quantum Efficiency Versus Wavelength of the Light Total quantum efficiency η(λ) in [%] for monochrome light versus wavelength of the light in [nm]. Not measured! 1 Basler ral8-8km

13 .1 Basic Parameters.1. Temporal Dark Noise Standard deviation of the temporal dark noise σ d time zero in [ e ]. referenced to electrons for exposure 'ral8-8km' (1 cameras), Std. Dev. Temporal Dark Noise Std. Dev. Temporal Dark Noise [e-] Camera 1 'ral8-8km' (1 cameras), Std. Dev. Temporal Dark Noise Histogram 8 Number Std. Dev. Temporal Dark Noise [e-] Figure 3: Temporal Dark Noise Item Symbol Typ. Std. Dev. Unit Remarks Temporal Dark Noise σ d 9.1 e Table 6: Temporal Dark Noise Basler ral8-8km 13

14 .1 Basic Parameters.1.3 Dark Current Dark current N d3 for a housing temperature of 3 C in [e /s]. Not measured!.1. Doubling Temperature Doubling temperature k d of the dark current in [ C]. Not measured! 1 Basler ral8-8km

15 .1 Basic Parameters.1.5 Inverse of Overall System Gain Inverse of overall system gain 1 K in [ e DN ]. Inverse of Overall System Gain [e-/dn] 'ral8-8km' (1 cameras), Inverse of Overall System gain 1 3 Camera Number 'ral8-8km' (1 cameras), Inverse of Overall System Gain Histogram Inverse of Overall System Gain [e-/dn] Figure : Inverse of Overall System Gain Item Symbol Typ. Std. Dev. Unit Remarks Inverse of Overall System Gain 1 K 6..5 Table 7: Inverse of Overall System Gain e DN Basler ral8-8km 15

16 .1 Basic Parameters.1.6 Inverse Photon Transfer 1 Inverse photon transfer in [ ] p ηk DN. 'ral8-8km' (1 cameras), Inverse Photon Transfer Inverse Photon Transfer [p~/dn] Camera Number 'ral8-8km' (1 cameras), Inverse Photon Transfer Histogram Inverse Photon Transfer [e-/dn] Figure 5: Inverse Photon Transfer Item Symbol Typ. Std. Dev. Unit Remarks Inverse Photon Transfer 1 ηk 1.8 TBD Table 8: Inverse Photon Transfer p DN λ = 55 nm 1 The main error in the inverse photon transfer is related to the error in the measurement of the illumination as described in section ηk Basler ral8-8km

17 .1 Basic Parameters.1.7 Saturation Capacity Saturation capacity µ e.sat referenced to electrons in [ e ]. 3 'ral8-8km' (1 cameras), Saturation Capacity Saturation Capacity [e-] Camera 1 'ral8-8km' (1 cameras), Saturation Capacity Histogram 8 Number Saturation Capacity [e-] Figure 6: Saturation Capacity Item Symbol Typ. Std. Dev. Unit Remarks Saturation Capacity µ e.sat 9 99 e Table 9: Saturation Capacity Basler ral8-8km 17

18 . Derived Data. Derived Data..1 Absolute Sensitivity Threshold Absolute sensitivity threshold µ p.min (λ) in [ p ] for monochrome light versus wavelength of the light in [nm]. µ p.min = σ d (1) η 'ral8-8km' (1 cameras), Absolute Sensitivity Threshold Absolute Sensitivity Threshold [p~] Camera 'ral8-8km' (1 cameras), Absolute Sensitivity Threshold Histogram 15 Number Absolute Sensitivity Threshold [p~] Figure 7: Absolute Sensitivity Threshold Item Symbol Typ. Std. Dev. Unit Remarks Absolute Sensitivity Threshold µ p.min 1 TBD p λ = 55 nm Table 1: Absolute Sensitivity Threshold 18 Basler ral8-8km

19 . Derived Data.. Signal-to-noise Ratio Signal-to-noise ratio SNR y (µ p ) is plotted versus number of photons µ p collected in a pixel during exposure time in [p ] for monochrome light with the wavelength λ given in [ nm]. The wavelength should be near the maximum of the quantum efficiency. A : SNR y = µ y µ y.dark σ y () B : SNR y = ηµ p (ηµp + σ d ) (3) Figure 8 shows the signal-to-noise ratio SNR y for monochrome light with the wavelength λ = 55 nm. 8 'ral8-8km' (1 cameras), SNR SNR [bit] 6 A B Mean Photon [bit] Figure 8: Signal-to-noise Ratio The maximum achievable image quality is given as SNR y.max. SNR y.max = µ e.sat () SNR y.max.bit = ld SNR y.max = log SNR y.max log (5) SNR y.max.db = log SNR y.max 6. SNR y.max.bit (6) Basler ral8-8km 19

20 . Derived Data 'ral8-8km' (1 cameras), SNR SNR Mean Photon [Photons/pixel] A B Figure 9: Signal-to-noise Ratio Item Symbol Typ. Std. Dev. Unit Remarks Maximum achievable SNR [bit] SNR y.max.bit bit Table 11: Maximum achievable SNR [bit] Item Symbol Typ. Std. Dev. Unit Remarks Maximum achievable SNR [db] SNR y.max.db..17 db Table 1: Maximum achievable SNR [db] Basler ral8-8km

21 . Derived Data..3 Dynamic Range Dynamic range DYN out.bit in [ bit]. DYN out = µ e.sat σ d (7) DYN out.bit = log (DYN out ) (8) 'ral8-8km' (1 cameras), Output Dynamic Range 1 Output Dynamic Range [bit] Camera 'ral8-8km' (1 cameras), Output Dynamic Range Histogram 1 8 Number Output Dynamic Range [bit] Figure 1: Output Dynamic Range Item Symbol Typ. Std. Dev. Unit Remarks Output Dynamic Range DYN out.bit bit Table 13: Output Dynamic Range Basler ral8-8km 1

22 .3 Raw Measurement Data.3 Raw Measurement Data.3.1 Mean Gray Value Mean gray value µ y (µ p ) in [DN] is plotted versus number of photons µ p in [p ] collected in a pixel during exposure time. Mean Gray Value Bright [DN] 'ral8-8km' (1 cameras), Mean Gray Value Bright 6 Mean Photon [Photons/pixel] Figure 11: Mean Gray Values of the Cameras with Illuminated Pixels Basler ral8-8km

23 .3 Raw Measurement Data.3. Variance of the Temporal Distribution of Gray Values The variance of the temporal distribution of gray values σ y.temp(µ p ) in [DN ] is plotted versus number of photons µ p in [p ] collected in a pixel during exposure time. Variance Gray Value Bright [DN^] 'ral8-8km' (1 cameras), Variance Gray Value Bright 6 Mean Photon [Photons/pixel] Figure 1: Variance Values for the Temporal Distribution of Gray Values with Illuminated Pixels Saturation Capacity The saturation point is defined as the maximum of the curve in figure 1. The abscissa of the maximum point is the number of photons µ p.sat where the camera saturates. The saturation capacity µ e.sat in electrons is computed according to the mathematical model as: µ e.sat = ηµ p.sat (9) Basler ral8-8km 3

24 .3 Raw Measurement Data.3.3 Mean of the Gray Values Dark Signal Mean of the gray values dark signal µ y.dark (T exp ) in [DN] time in [s]. is plotted versus exposure 'ral8-8km' (1 cameras), Mean Gray Value Dark 35 Mean Gray Value Dark [DN] Exposure Time [ms] Figure 13: Mean Gray Values for the Cameras in Darkness Basler ral8-8km

25 .3 Raw Measurement Data.3. Variance of the Gray Value Temporal Distribution in Darkness The variance of the temporal distribution of gray values in darkness σ y.temp.dark(t exp ) in [DN ] is plotted versus exposure time T exp in [s]. 'ral8-8km' (1 cameras), Variance Gray Value Dark Variance Gray Value Dark [DN^] Exposure Time [ms] Figure 1: Variance Values for the Temporal Distribution of Gray Values in Darkness Temporal Dark Noise The dark noise for exposure time zero is found as the offset of the linear correspondence in figure 1. Match a line (with offset) to the linear part of the data in the diagram. The dark noise for exposure time zero σd is found as the offset of the line divided by the square of the overall system gain K. σ d = σ y.temp.dark (T exp = ) K (1) Basler ral8-8km 5

26 .3 Raw Measurement Data.3.5 Light Induced Variance of the Temporal Distribution of Gray Values The light induced variance of the temporal distribution of gray values in [DN ] is plotted versus light induced mean gray value in [DN]. Variance Gray Value (Bright - Dark) [DN^] 'ral8-8km' (1 cameras), Diff. Variance vs Diff. Mean Gray Value Mean Gray Value (Bright - Dark) [DN] Figure 15: Light Induced Variance of the Temporal Distribution of Gray Values Versus Light Induced Mean Gray Value The overall system gain K is computed according to the math- Overall System Gain ematical model as: K = σ y.temp σ y.temp.dark µ y µ y.dark (11) which describes the linear correspondence in figure 15. Match a line starting at the origin to the linear part of the data in this diagram. The slope of this line is the overall system gain K. 6 Basler ral8-8km

27 .3 Raw Measurement Data.3.6 Light Induced Mean Gray Value The light induced mean gray value µ y µ y.dark in [ DN] is plotted versus the number of photons collected in a pixel during exposure time Kµ p in [ p ]. Mean Gray Value (Bright - Dark) [DN] 'ral8-8km' (1 cameras), Difference Mean Gray Value 1 3 Mean Photon [Photons/pixel] Figure 16: Light Induced Mean Gray Value Versus the Number of Photons Total Quantum Efficiency The total quantum efficiency η is computed according to the mathematical model as: η = µ y µ y.dark (1) Kµ p which describes the linear correspondence in figure 16. Match a line starting at the origin to the linear part of the data in this diagram. The slope of this line divided by the overall system gain K yields the total quantum efficiency η. The number of photons µ p is calculated using the model for monochrome light. The number of photons Φ p collected in the geometric pixel per unit exposure time [p /s] is given by: Φ p = EAλ (13) hc with the irradiance E on the sensor surface [W/m ], the area A of the (geometrical) pixel [m ], the wavelength λ of light [m], the Planck s constant h Js, and the speed of light c m/s. The number of photons can be calculated by: µ p = Φ p T exp (1) during the exposure time T exp. Using equation 1 and the number of photons µ p, the total quantum efficiency η can be calculated as: η = hc 1 AT exp E 1 λ µ p µ y.dark. (15) K Basler ral8-8km 7

28 .3 Raw Measurement Data.3.7 Dark Current Versus Housing Temperature The logarithm to the base of the dark current in [e /s] versus deviation of the housing temperature from 3 C in [ C] Not measured! 8 Basler ral8-8km

29 5 Characterizing Total and Spatial Noise 5 Characterizing Total and Spatial Noise 5.1 Basic Parameters Spatial Offset Noise Standard deviation of the spatial offset noise σ o referenced to electrons in [ e ]. 5 'ral8-8km' (1 cameras), DSNU188 DSNU188 [e-] Camera 1 'ral8-8km' (1 cameras), DSNU188 Histogram 1 Number DSNU188 [e-] Figure 17: Spatial Offset Noise ( DSNU 188 ) Item Symbol Typ. Std. Dev. Unit Remarks Spatial Offset Noise ( DSNU 188 ) σ o.3.1 e Table 1: Spatial Offset Noise ( DSNU 188 ) Basler ral8-8km 9

30 5.1 Basic Parameters 5.1. Spatial Gain Noise Standard deviation of the spatial gain noise S g in [ %]..3 'ral8-8km' (1 cameras), PRNU188.5 PRNU188 [%] Camera 3 'ral8-8km' (1 cameras), PRNU188 Histogram 5 Number PRNU188 [%] Figure 18: Spatial Gain Noise ( PRNU 188 ) Item Symbol Typ. Std. Dev. Unit Remarks Spatial Gain Noise ( PRNU 188 ) S g.. % Table 15: Spatial Gain Noise ( PRNU 188 ) 3 Basler ral8-8km

31 5. Raw Measurement Data 5. Raw Measurement Data 5..1 Standard Deviation of the Spatial Dark Noise Standard deviation of the spatial dark noise in [DN] versus exposure time in [s]. Spatial Std. Dev. Gray Value Dark [DN] 'ral8-8km' (1 cameras), Spatial Std. Dev. Gray Value Dark Exposure Time [ms] Figure 19: Standard Deviation of the Spatial Dark Noise From the mathematical model, it follows that the variance of the spatial offset noise σ o should be constant and not dependent on the exposure time. Check that the data in the figure 19 forms a flat line. Compute the mean of the values in the diagram. The mean divided by the conversion gain K gives the standard deviation of the spatial offset noise σ o. DSNU 188 = σ o = σ y.spat.dark K The square of the result equals the variance of the spatial offset noise σo. (16) Basler ral8-8km 31

32 5. Raw Measurement Data 5.. Light Induced Standard Deviation of the Spatial Noise Light induced standard deviation of the spatial noise in [DN] versus light induced mean of gray values [DN]. Std. Dev. Gray Value (Bright - Dark) [DN] 'ral8-8km' (1 cameras), Spatial Gain Noise Mean Gray Value (Bright - Dark) [DN] Figure : Light Induced Standard Deviation of the Spatial Noise The variance coefficient of the spatial gain noise Sg or its standard deviation value S g respectively, is computed according to the mathematical model as: PRNU 188 = S g = σ y.spat σ y.spat.dark µ y µ y.dark, (17) which describes the linear correspondence in figure. Match a line through the origin to the linear part of the data. The line s slope equals the standard deviation value of the spatial gain noise S g. 3 Basler ral8-8km

33 REFERENCES References [1] EUROPEAN MACHINE VISION ASSOCIATION (EMVA): EMVA Standard Standard for Characterization and Presentation of Specification Data for Image Sensors and Cameras (Release A1.3). 6 Basler ral8-8km 33

34 Basler AG An der Strusbek 6-6 D-96 Ahrensburg To whom it may concern Appendix Measurement protocol - Basler racer Camera Series 1) Deviation in Quantum Efficiency ) Photo Response Non Uniformity Ahrensburg, October, 13 phone: fax: marc.nehmke@baslerweb.com Dear Sir or Madam, Please note that the current EMVA Standard 188 camera calculation model does not fit exactly the most recent technological advance as e.g. realized in the latest CMOS sensors. The quantum efficiency information given in this camera measurement protocol was calculated assuming the current EMVA Standard 188 regulations. However, the quantum efficiency information (see below) reported by the sensor manufacturer for the sensors used in the Basler racer camera series differs from the quantum efficiency information given in this camera specification.

35 Furthermore, please note that the Photo Response Non Uniformity (PRNU) information given in this report refers to the use of shading compensation for a specific operating point. Independent evaluations of series cameras using the cameras default shading settings confirmed a PRNU of.5% on average. Best regards Marc Oliver Nehmke, Product Manager