Measurement of luminance distributions in the field of ... · Measurement of luminance...

Measurement of luminance distributions in the field of view for glare evaluations

Jan Wienold & Peter HansenInterdisciplinary Laboratory of Performance-Integrated Design (LIPID)Ecole polytechnique fédérale de Lausanne (EPFL)

WORKSHOPMeasurement of luminance distributions in the field of view for glare evaluations

1 | PH | 10.2017

HDR Cameras

2 | PH | 23-10.2017

Reliable glare studies

Reliable luminance measurements

2 different methods• Absolute calibration• ”Self-calibration”

• Automatic algorithm

Glare and luminance

3 | PH | 23-10.2017

Absolute HDR camera calibration

0 15 30 45 60 75 90 105 120 135 150 165 180 195 210 225 240 255

Pixel value

Response curve

The shape of the response curve

4 | PH | 23-10.2017

Absolute HDR camera calibration

00,10,20,30,40,50,60,70,80,9

0 30 60 90 120 150 180 210 240

Pixel value

Exposure 1

00,10,20,30,40,50,60,70,80,9

0 30 60 90 120 150 180 210 240

Pixel value

Exposure 2

00,10,20,30,40,50,60,70,80,9

0 30 60 90 120 150 180 210 240

Pixel value

Exposure ...

00,10,20,30,40,50,60,70,80,9

0 30 60 90 120 150 180 210 240

Pixel value

Exposure ...

For each aperture know at least one luminance value

5 | PH | 23-10.2017

Luminance values calculated based on pixel value, response curve and exposure

HDR generation

6 | PH | 23-10.2017

Automatic algorithm HDR camera calibration

0 15 30 45 60 75 90 105 120 135 150 165 180 195 210 225 240 255

Pixel value

Response curve

20 < 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑉𝑉𝑉𝑉𝑃𝑃𝑉𝑉𝑃𝑃 < 200

An automatic calibration algorithm is used to derive a response curve

7 | PH | 23-10.2017

Different programs• PFSTools• hdrgen

• Photosphere• WebHDR

Use up-to-date versions – also of Radiance (current version 5.1)

Automatic algorithm HDR camera calibration

8 | PH | 23-10.2017

DSLR camera “self/calibration” procedure

Create LDR images

Create HDR image

Cut images

Calibrate for vignetting

Calibrate for projection

Luminance calibration

Illuminance check

9 | PH | 23-10.2017

General camera settings

ISO speed 100White balance DaylightImage quality .JPEG (large)Color space sRGBFocus Infinity (auto off)Picture style Standard

10 | PH | 23-10.2017

• Auto bracketing• Computer control

• Manufactures• qDSLRdashboard

General camera setup

11 | PH | 23-10.2017

1 EV stop between images

12 | PH | 23-10.2017

• Avoid movement in scene (avoid ghosting)• Avoid shaking camera• Use a tripod• Use a computer for bracketing • Black to white image series• Use targets in the scene

• High luminance• Low luminance

• Measure vertical illuminance

13 | PH | 23-10.2017

Create LDR images

Create HDR image

Cut images

Illuminance check

14 | PH | 23-10.2017

Generate HDR hgrgen –o –r –a –e –f –g –x

‒ o = output file‒ r = response curve file‒ a = automatic exposure alignment‒ e = exposure adjustment ‒ f = flare removal‒ g = ghost removal‒ x = over and under exposure removal

HDR generation

15 | PH | 23-10.2017

HDR generation

1: 89.809 cd/m2

2: 291 cd/m21: 2.653.980 cd/m2

2: 281 cd/m2

With –x option

Konica Minolta: 1: 1.674.000 cd/m2

2: 275 cd/m2

Without –x option

16 | PH | 23-10.2017

Check LDR images• Darkest should not have pixels > 200• Whitest should not have pixels < 20

HDR generation

17 | PH | 23-10.2017

Response curve

Scene should contain• bright areas• dark areas• white and/or grey areas

18 | PH | 23-10.2017

hgrgen –o –r –a –e –f –g to generate HDR and curve

With –r option give name to response curve fileSave it and use it to make HDR images

[n] degree polynomial

Response curve – HDR generation

19 | PH | 23-10.2017

Response curve

Differs for all cameras and apertures

20 | PH | 23-10.2017

Create LDR images

Create HDR image

Cut images

Illuminance check

21 | PH | 23-10.2017

Cut HDR image

pcompos -x [x-max] -y [y-max] -h [org_image.hdr] -[x-min] –[y-min] > [cut_image.hdr]

x-maxx-miny-max

22 | PH | 23-10.2017

Create LDR images

Create HDR image

Cut images

Illuminance check

23 | PH | 23-10.2017

Vignetting

-90 -80 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90

24 | PH | 23-10.2017

Vignetting

Normal camera rotation point

25 | PH | 23-10.2017

Vignetting

Nodal rotation point

26 | PH | 23-10.2017

Vignetting

Second setup images curtesy of Clotilde Pierson

27 | PH | 23-10.2017

Vignetting

-90 -80 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90Re

Aperture – f/11

-90 -80 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90

Aperture – f/3.2

Normalized to lens center

28 | PH | 23-10.2017

Vignetting

-90 -70 -50 -30 -10 10 30 50 70 90

All apertures

2,8 3,2 3,5 4 4,5 5 5,6 6,3 7,1 8

9 10 11 13 14 16 18 20 22

29 | PH | 23-10.2017

Create a .cal file with the polynomial equation

Vignetting

y = -3,771557x6 + 2,658279x5 + 6,604217x4 - 8,870208x3 + 3,430605x2 - 0,474940x + 1,001537R² = 0,995700

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

Relative angle

Aperture –f/3.2

Establish 6th order polynomial for vignetting curve

30 | PH | 23-10.2017

Apply to images with

pcomb –f [path_to/vignetting.cal] –o [original_image.hdr] > [corrected_image.hdr]

VignettingCal file example

31 | PH | 23-10.2017

Vignetting

• Canon 70D • Sigma 4.5mm f/2.8 fisheye lens

32 | PH | 23-10.2017

Create LDR images

Create HDR image

Cut images

Illuminance check

33 | PH | 23-10.2017

Projection

34 | PH | 23-10.2017

Projection

180 px133 px

35 | PH | 23-10.2017

From Sigma:

Projection

36 | PH | 23-10.2017

Fisheye correction file from: https://www.radiance-online.org/cgi-bin/viewcvs.cgi/ray/src/cal/cal/

Projection

fisheye_corr.cal

pcomb –f [path_to/fisheye_corr.cal] -o [original_image.hdr] > [corrected_image.hdr]

37 | PH | 23-10.2017

Also make sure the fisheye lens fits the sensor frame

Projection

Sensor frame

38 | PH | 23-10.2017

Also make sure the fisheye lens fits the sensor frame

Projection

Sensor frame

39 | PH | 23-10.2017

Create LDR images

Create HDR image

Cut images

Illuminance check

40 | PH | 23-10.2017

Luminance adjustment

Target Konica Image Factor1 151 140,2 1,0772 268 247,2 1,0843 107 98,5 1,086

pcomb -s 1.08 -o [original_image.hdr] > [adjusted_image.hdr]

Should be less than ± 20 %

41 | PH | 23-10.2017

Luminance adjustment

Target Konica Image Factor1 151 151,1 0,9992 268 266,2 1,0073 107 107,8 0,993

pcomb -s 1.08 -o [original_image.hdr] > [adjusted_image.hdr]

Should be less than ± 20 %

42 | PH | 23-10.2017

Check header

Remove or reset exposure to 1

Set “View” string to the right values (no “tab” in front)-vta –vh 180 –vv 180

Remember to use the current version of Radiance (5.1 at the moment)

43 | PH | 23-10.2017

Create LDR images

Create HDR image

Cut images

Illuminance check

44 | PH | 23-10.2017

Measure illuminance in scene

Calculate from HDR image

Illuminance check

LMT Image Factor1150 1145,8 1,004

45 | PH | 23-10.2017

Do NOT use this to adjust luminance

This is a check that the “energy balance” of the image is correct

± 10 % is okay but should be random

Illuminance check

46 | PH | 23-10.2017

High luminance scenes

47 | PH | 23-10.2017

Be careful - long exposures with very high luminance can “burn out” the image sensor

48 | PH | 23-10.2017

Use filters to reduce light on sensor

49 | PH | 23-10.2017

30 sec exposure at f/11

However, filters raises other problems

Grainy HDR

50 | PH | 23-10.2017

6 layers of 0,3 filters on window

51 | PH | 23-10.2017

1: 1.651.000 cd/m2

2: 274 cd/m21: 762.973 cd/m2

2: 12.110 cd/m21: 2.653.980 cd/m2

2: 281 cd/m2

Canon 0.1 % filter TechnoTeam 0.01 % filter Canon no filter

12 2 12 1

Konica Minolta: 1: 1.674.000 cd/m2

2: 275 cd/m2

52 | PH | 23-10.2017

1: 264.000 cd/m2

2: 327 cd/m21: 342.192 cd/m2

2: 28.125 cd/m21: 388.479 cd/m2

2: 350 cd/m2

Canon 0.1 % filter TechnoTeam 0.01 % filter Canon no filter

12 2 12 1

Konica Minolta: 1: 261.000 cd/m2

2: 313 cd/m2

53 | PH | 23-10.2017

Avoid very high luminancesCheck that the recorded values are correct

54 | PH | 23-10.2017

Avoid flicker• Influences pixel values

R: 148G: 162B: 189

R: 159G: 181B: 203

55 | PH | 23-10.2017

When “self-calibrating”• Use recommended camera settings• Use a tripod• Create good response curves• Correct for vignetting• Correct for projection• Luminance adjust • Check vertical illuminance• Check the header• Be careful with very high luminance

Use the forums; Radiance HDRI is a good starthttps://www.radiance-online.org/community/mailing-lists

Synthesis

Measurement of luminance distributions in the field of ... · Measurement of luminance...

Documents

Tone mapping con Luminance

TOPCON Luminance Meter_BM9-Manual

2015 Luminance LED Catalog

Automatic Luminance Control

Seem - focalpointlights.com · luminance data ) ® ® luminance data )

Luminance field of the façades - stars4all.eu · façades, and also luminance contrast. This parameter (Luminance contrast - LC) it is not used in present for façades, but is well

Lloyds 2020 - Sinalux LLL Everlux LLL EN · 2020. 3. 23. · Fire Retardance: NBR 13434-3:2018 (Clause 4.1) Luminance Performance. Luminance Performance. Luminance Performance: Luminance

Luminance Notes

Automotive Biometric Automatic Luminance Control System

LUMIOS MkIII Luminance Photometer for Tunnels

Ramayana: The Thread of Luminance Work Flow

LIGHT & DISPLAY - KONICA MINOLTA Europe · Model Colour Luminance Meter LS-150 Colour Luminance Meter LS-160 Measuring angle 1° 1/3° Luminance meter class (Applicable standard)

Interaction between luminance gratings and disparity gratingspsychology.nottingham.ac.uk/research/vision/papers/Wright... · 2008-09-08 · Interaction between luminance gratings

Neural basis of location-specific pupil luminance modulation · luminance decrease (3) (referred to as global luminance modu-lation) to regulate the trade-off between image acuity

Relationship between thermal and luminance distributions in high-power lateral GaN/InGaN light-emitting diodes D.P. Han, J.I. Shim and D.S. Shin ELECTRONICS

Hue Saturation & Luminance

Preference for luminance histogram regularities in natural ... · Thus, if we tend to like shiny and/or fresh things, which tend to gen-erate higher skew in luminance distributions,

ANGULAR DISTRIBUTIONS, ENERGY DISTRIBUTIONS- …

Relative luminance and figure-background segmentation

Ron Yanovich & Guy Peled 1. Contents Grayscale coloring background Luminance / Luminance channel Segmentation Discrete Cosine Transform K-nearest-neighbor