Daytime veiling glare in automobiles
caused by dashboard reflectance
by
Andreas Dunsäter & Marcus Andersson
Division of Industrial Ergonomics
Department of Management and Engineering
Master thesis LIU-IEI-TEK-A--08/00376--SE
i
Preface
This thesis work has been performed at the Department of Management and Engineering (IEI) at Linköping University of Technology, Sweden. The work was commissioned by Saab Automobile AB at the department of Human Vehicle Integration (HVI). The work started in October 2007 and was finished in April 2008. We have a lot of people to thank for helping us to carry out this work in a good way.
First of all we like to thank our examinator and supervisor at the University and our supervisor at Saab. Thank you Torbjörn Alm, for your help to guide us through this work and your valuable help with this report. Another big thank you to Claes Edgren for your help and big enthusiasm about this work. Thank you Bo Magnusson for your help.
We also like to express our gratitude to: Anne Johansson
Magnus Olsson Hillevi Hemphälä Fredrich Claezon Anette Karltun
ii
Abstract
Veiling glare has always existed in cars, but during the last years it has been brought up as a big problem. One reason is that glossier materials are being used in car interior design. Another reason is that the customers who buy the cars are getting more quality conscious. They demand to get top quality for the high price that they pay for a car, and veiling glare problems could be regarded as “low quality”.
Veiling glare is when light hits the car interior and reflects into the windshield, causing mirror-like images in the windshield (ghost images). This can impair the driving experience in two ways. It can lower the contrast of the road scene and it may be a cluttering for the driver.
This work handles daytime veiling glare from dashboard reflectance. The purpose was to investigate the area and to see if Saab can avoid the problem with veiling glare by using virtual prototyping (see chapter 3.3.1). This has been done by examining if the light simulation software Speos can be used to simulate and predict veiling glare, and thereby be used as a tool for better design.
iii
Sammanfattning
Reflexer från interiören i framrutan har alltid existerat i bilar, men under de senaste åren har det blivit ett allt större problem. En anledning till det är att den interiöra designen blir mer och mer avancerad och att glansigare material används. En annan anledning är att bilkunderna har blivit mer kvalitetsmedvetna och bara accepterar toppkvalité av en så dyr produkt, och reflexer i framrutan kan uppfattas som låg kvalité.
Problemet med reflektionerna uppstår när ljus träffar interiören i bilen och reflekteras upp i framrutan. Det gör att förarens körkvalité försämras på två sätt. Dels kan det sänka kontrasterna i förarens synfält och dels kan det vara en störande faktor.
Det här arbetet har begränsats till att hantera reflektioner i framrutan från instrumentpanelen under dagtid. Syftet med arbetet var att undersöka om Saab kunde undvika problemen med reflektionerna genom att använda sig av ”virtual prototyping” (se kapitel 3.3.1). Detta har gjorts genom att undersöka om den ljussimulerande mjukvaran Speos kan användas för att förutsäga reflektioner från interiören i framrutan, och kan därmed användas som ett verktyg för bättre design.
iv
Table of Contents
1. INTRODUCTION ... 1 1.1 BACKGROUND ... 1 1.1.1 Saab Automobile AB ... 2 1.1.2 Optis ... 2 1.1.3 Problem Statement ... 2 1.2 SCOPE ... 2 2. PURPOSE AND RESEARCH QUESTIONS ... 3 3. THEORETICAL FRAME OF REFERENCE ... 4 3.1 LIGHT ... 4 3.1.1 Illuminance ... 6 3.1.2 Luminance ... 6 3.1.3 To Measure Luminance and Illuminance ... 6 Calculating ... 7 Measuring ... 7 Black body absorber ... 8 Luminance Contrast Ratio ... 9 3.1.4 Contrast ... 9 3.1.5 Reflection ... 10 3.1.6 Refraction ... 12 3.2 THE HUMAN EYE ... 12 3.2.1 Glare ... 13 Discomfort ... 13 Disability ... 13 3.2.2 Adaptation ... 13 3.2.3 Phototropism ... 133.3 VISUAL REDUCTION WHILE DRIVING ... 14
3.3.1 Veiling Glare ... 15 3.4 METHODOLOGY ... 17 3.4.1 Simulation‐based design ... 17 3.4.2 Methods for data collection ... 17 Method of Adjustment ... 18 Staircase method ... 18 Double Staircase Method ... 20 Questionnaire ... 20 Objective measures ... 20 4. METHOD ... 21 4.1 PROJECT REALIZATION ... 21 4.2 STARTUP ... 22 4.3 STAIRCASE METHOD ... 23 4.4 QUESTIONNAIRE ... 23 5. APPARATUS ... 24 5.1 SPEOS ... 24 5.1.1 Squale ... 24 BRDF ... 25
v 5.1.2 Ray Tracing... 25 5.1.3 Photon Mapping ... 25 5.1.4 Viewer ... 25 5.1.5 CIE ... 26 5.2 SMART EYE ... 26 5.3 DRIVING SIMULATOR ... 27 5.4 PHOTOMETER ... 28 5.5 LIGHT TRAP ... 28 6. REALIZATION ... 29
6.1 DIVIDING THE WINDSHIELD INTO ZONES ... 29
6.2 ANALYZE OF SMART EYE DATA ... 30
6.3 SPEOS ... 31 6.3.1 Scanning of Materials ... 31 6.3.2 Modifying the CAD‐model ... 32 6.3.3 Options in Speos ... 34 6.3.4 Viewer ... 36 6.3.5 Simulated images with Speos ... 37 6.4 DRIVER STUDY ... 39 6.4.1 Driver Study Procedure ... 39 Preparation for Driver study ... 39 Realization of the Driver Study ... 40 6.4.2 Indoors ... 42 6.4.3 Outdoors ... 43 7. RESULTS ... 44 7.1 VALIDATING SPEOS ... 44 7.1.1 Comparing Luminance Values ... 44 7.1.2 Comparing Photorealism ... 46 7.2 DRIVER STUDY ... 47 7.2.1 Intensity test ... 47 7.2.2 Position test ... 48 8. CONCLUSIONS AND FUTURE WORK ... 51 9. DISCUSSION ... 55 9.1 STARTUP ... 55 9.2 OPTIS ... 55 9.2.1 Validating Speos ... 55 9.2.2 Comparing Luminance Values ... 56 9.2.3 Comparing Photorealism ... 56 9.3 DRIVER STUDY ... 56 9.3.1 Intensity test ... 56 9.3.2 Position test ... 57 10. BIBLIOGRAPHY ... 59 APPENDIX A ... 62 APPENDIX B ... 68
vi APPENDIX C ... 69 APPENDIX D ... 70 APPENDIX E ... 75 APPENDIX F ... 81 APPENDIX G ... 83 APPENDIX H ... 84
vii
Table of Figures
FIGURE 1‐ ELECTROMAGNETIC SPECTRUM (BRITANNICA, 2008) ... 4
FIGURE 2 – CIE SENSITIVITY STANDARD OF PHOTOPIC AND SCOTOPIC VISION (MODIFIED IMAGE) (WIKIPEDIA, N.D.). ... 5
FIGURE 3 – SENSITIVITY DURING PHOTOPIC AND SCOTOPIC VISION (STARBY, 1992)... 5
FIGURE 4 – RELATIONSHIP BETWEEN LUMINANCE AND ILLUMINANCE (STARBY, 1992) ... 6
FIGURE 5 – REFLECTION FACTORS (STARBY, 1992) ... 7
FIGURE 6 – EXAMPLE OF SAME LUMINANCE (STARBY, 1992) ... 8
FIGURE 7 – CONTRAST RELATIONSHIP BETWEEN SURFACES (STARBY, 1992) ... 9
FIGURE 8 ‐ DIFFERENT CONTRASTS ... 9
FIGURE 9 – RELATIVE CONTRAST SENSITIVITY ACCORDING TO CIE (KELLEY, JONES, & GERMER, 2008) ... 10
FIGURE 10 – EXPERIMENT TO DEFINE DIFFERENT REFLECTIONS (KELLEY, JONES, & GERMER, 2008) ... 11
FIGURE 11 – BRDF EXPERIMENT (KELLEY, JONES, & GERMER, 2008)... 11
FIGURE 12 – THREE DIFFERENT REFLECTION COMPONENTS (KELLEY, JONES, & GERMER, 2008) ... 11
FIGURE 13 – REFRACTION (KELLEY, JONES, & GERMER, 2008) ... 12
FIGURE 14 ‐ THE HUMAN EYE (LIGHT: HUMAN EYE, 2007) ... 12
FIGURE 15 ‐ VEILING LUMINANCE DISTRIBUTION (MEFFORD ET AL., 2003) ... 14
FIGURE 16 – (LEFT) EXAMPLE OF CONTRAST LOWERING VEILING GLARE. (RIGHT) EXAMPLE OF GHOST IMAGES. ... 15
FIGURE 17 ‐ REFLECTANCE DEPENDING ON WINDSHIELD RAKE ANGLE (SCHUMANN & FLANNAGAN, 1997) ... 16
FIGURE 18 ‐ STAIRCASE METHOD (CORNSWEET, 1962) ... 19
FIGURE 19 ‐ DOUBLE STAIRCASE METHOD (CORNSWEET, 1962) ... 20
FIGURE 20 ‐ AN OVERVIEW FIGURE OF THE REALIZATION OF THE THESIS WORK ... 22
FIGURE 21 – SCREENSHOT FROM SPEOS (OPTIS WORLD, 2008) ... 24
FIGURE 22 ‐ SQUALE TOOL (SQUALE, 2008) ... 24
FIGURE 23 ‐ EXAMPLES OF CIE SKY MODELS (DAYLIGHTING, 2007) ... 26
FIGURE 24 ‐ CAMERA AND IR‐FLASH ILLUMINATORS (TECHNOLOGY, 2008)... 27
FIGURE 25 ‐ HOW THE SMART EYE SYSTEM WORKS (TECHNOLOGY, 2008) ... 27
FIGURE 26 – HAGNER S3 (PERSSON, PHOTAC, HAGNER, & AB, N.D.) ... 28
FIGURE 27 – DEFINED AREA WHICH LIGHT SHOULD BE AVOIDED (STARBY, 1992) ... 29
FIGURE 28 ‐ THE WINDSHIELD ZONES ... 30
FIGURE 29 ‐ THE VIEWING ZONES ON THE DASHBOARD ... 30
FIGURE 30 ‐ GAZE POINTS FROM ONE OF THE SMART EYE FILES ... 31
FIGURE 31 – MATERIAL LIBRARY IN CATIA ... 32
FIGURE 32 – DESIGN LINES IN CATIA ... 33
FIGURE 33 – CHROME LIST ... 33
FIGURE 34 – HDR IMAGE ... 35
FIGURE 35 – VIEWER SOFTWARE SHOWING THE PHOTOMETRIC IMAGE OF THE SIMULATION ... 36
FIGURE 36 ‐ LUMINANCE VALUES FROM FIGURE 36 ... 36
FIGURE 37 – LIGHT MATERIALS AGAINST A DARK DASHBOARD BACKGROUND ... 37
FIGURE 38 – DARK DETAIL AGAINST A LIGHT DASHBOARD BACKGROUND ... 38
FIGURE 39 ‐ SIMULATED IMAGE OF DARK LINES ON LIGHT BACKGROUND ... 38
FIGURE 40 ‐ LUMINANCE VALUES FROM FIGURE 39 ... 39
FIGURE 41 – PIECE OF PAPERS THAT WAS USED DURING THE INTENSITY TEST ... 40
FIGURE 42 ‐ FIELD OF VISION DURING THE INTENSITY TEST ... 41
FIGURE 43 – FIELD OF VISION DURING THE POSITIONING TEST ... 41
FIGURE 44 – SIMULATOR AT LINKÖPING’S UNIVERSITY WITH THE LAMP MOUNTED ABOVE THE COCKPIT ... 42
FIGURE 45 ‐ COMPARISON BETWEEN OUR MEASUREMENTS AND THE RESULT FROM THE SIMULATIONS IN SPEOS ... 44
FIGURE 46 – COMPARISON BETWEEN OUR MEASUREMENTS AND THE RESULT FROM SPEOS ... 45
FIGURE 47 ‐ (LEFT) SIMULATED IMAGE OF LIGHT MEDIUM SIZED LINES ON DARK BACKGROUND. (RIGHT) PHOTOGRAPH OF THE SAME SCENARIO. ... 46
FIGURE 48 ‐ (LEFT) SIMULATED IMAGE OF DARK MEDIUM SIZED LINES ON LIGHT BACKGROUND. (RIGHT) PHOTOGRAPH OF THE SAME SCENARIO. ... 46
FIGURE 49 ‐ RESULTS FROM DRIVER STUDY MADE IN THE CAR DRIVING SIMULATOR ... 47
FIGURE 50 ‐ RESULTS FROM DRIVER STUDY MADE OUTDOOR IN REAL TRAFFIC ... 48
viii
FIGURE 52 ‐ RESULTS FROM THE SIMULATOR ... 49
FIGURE 53 – RESULTS FROM THE SIMULATOR COLLECTED IN A TABLE ... 49
FIGURE 54 ‐ RESULTS WHEN DRIVING OUTDOORS IN A REAL ENVIRONMENT ... 50
FIGURE 55 ‐ RESULTS FROM THE OUTDOOR TEST COLLECTED IN A TABLE ... 50
FIGURE 56 – RESULTS COLLECTED FROM THE SMART EYE ... 50
FIGURE 57 – VEILING GLARE IN BOTH MAIN WINDSHIELD AND THE SIDE WINDOW ... 52
FIGURE 58 – (ABOVE) CHROME LIST AGAINST LIGHT DASHBOARD BACKGROUND, (BELOW) CHROME LIST AGAINST DARK DASHBOARD
BACKGROUND ... FEL! BOKMÄRKET ÄR INTE DEFINIERAT.
FIGURE 59 – COMPARING DIFFERENT MATERIALS, VOLVO V70, AUDI A4 AND VOLVO C30 .... FEL! BOKMÄRKET ÄR INTE DEFINIERAT.
FIGURE 60 – LIGHT DETAILS ON A DARK DASHBOARD BACKGROUND, LIGHTENED FROM DIFFERENT DIRECTIONS ... FEL! BOKMÄRKET ÄR
INTE DEFINIERAT.
FIGURE 61 – VIEWPORTS FROM SHORT, MEDIUM AND A TALL PERSON ... FEL! BOKMÄRKET ÄR INTE DEFINIERAT.
FIGURE 62 – GLOSSY DETAILS ON A BLACK DASHBOARD BACKGROUND, LIGHTENED FROM DIFFERENT DIRECTIONS FEL! BOKMÄRKET ÄR
INTE DEFINIERAT.
FIGURE 63 – DIFFERENCE BETWEEN A BLACK AND BEIGE DASHBOARD ... FEL! BOKMÄRKET ÄR INTE DEFINIERAT.
FIGURE 64 – CHROME DETAIL TOWARDS A BLACK AND BEIGE DASHBOARD BACKGROUND ... FEL! BOKMÄRKET ÄR INTE DEFINIERAT.
FIGURE 65 – BLACK DETAIL TOWARDS A BLACK DASHBOARD BACKGROUND AND A BEIGE DETAIL TOWARDS A BEIGE DASHBOARD
BACKGROUND ... FEL! BOKMÄRKET ÄR INTE DEFINIERAT.
FIGURE 66 – DARK GLOSSY MATERIAL TOWARDS A MATT DASHBOARD BACKGROUND ... FEL! BOKMÄRKET ÄR INTE DEFINIERAT.
FIGURE 67 – TWO BLACK AND TWO BEIGE DASHBOARD MATERIALS WITH SMALL DIFFERENCE IN MATT SURFACES FEL! BOKMÄRKET ÄR
INTE DEFINIERAT.
FIGURE 68 – (ABOVE) LIGHT MATERIALS, LIGHT TEXTILE TOWARDS PLASTIC DASHBOARD BACKGROUND (BELOW) PLASTIC DETAIL
1
1. Introduction
More and more people are getting aware of the problem called veiling glare while driving, making the problem a big issue for the car manufacturers. Veiling glare is when the dashboard, or parts of the dashboard, is reflected into the windshield causing veiling images (ghost images) that disturbs and impairs the drivers contrast and vision of the road scene. This phenomenon can be very irritating for the driver and it can also be a safety hazard. Most people have not thought of it as a problem before, but when someone tells them about it, they get more aware of the problem.
Veiling glare in cars can be divided into two types of glare and problems. One problem is light colored dashboards that reflect into the windshield causing “a wall” that the driver has to look through to see the road scene. This veiling glare reduces all the contrasts of the road scene, making it harder to detect objects. This can both be disturbing for the driver and also become a safety hazard. The second kind of problem is contrast differences on the dashboard, automatically causing contrast differences in the windshield because of veiling glare, called a ghost image. The details that appear in the windshield can be very irritating for the driver. This problem could be compared with a cluttered screen in a computer-based environment. The driver study performed in this project, were focused on the level of acceptance with ghost images displayed in the windshield.
In this report we are handling the problem with veiling glare and how to predict it, in order to give a basis for better design.
1.1 Background
The phenomenon called veiling glare have always existed and is hard to completely avoid in windshields. The question is how it can be reduced so that it disturbs the driver as little as possible. The problem with veiling glare is growing since the car interior designers are using more chrome and other shiny materials in the design. The attractive shiny dashboard causes big problems with veiling glare. This has made pedantic customers use dull dashboard covers, conceal shiny details with dark tape and black spray light colored parts, to get rid of veiling glare. But it is not only the meticulous customers that complain. Costumers of today have higher expectations and quality demands on products that they have spent a lot of money on. It is also important to deliver the best car when the customer will compare different manufactures. The competition between the different car manufactures today is extremely high and it is important that the car has no malfunctions which the customer will regard as disturbing. To meet these demands and further improve the quality, the car manufacturers have begun to use simulation-based design and virtual prototyping.
It is possible to perform a study to catch the level of acceptance with veiling glare, which could be used when a virtual prototype is created in Speos. This way Saab can decide in an early state in the design process if something should be modified to improve the design. General Motors and Saab have not yet decided to buy the light simulation software Speos. To test the usefulness of the software they have used consulting favors from Optis and asked us to perform a research how useful the software could be for Saab.
2
1.1.1 Saab Automobile AB
This work has been done in cooperation with Saab Automobile AB, Trollhättan. The name Saab comes from “Svenska Aeroplan Aktiebolaget” which means “Swedish Aircraft Company”. The company was founded in Trollhättan north of Gothenburg in 1937, where they started producing military aircraft for the Swedish Air Force. In 1944 they expanded into civil aviation and in 1947 they started to produce cars. Today the car business part of the company, Saab Automobile AB, is owned by the world’s largest auto maker General Motors. A big part of Saab is still kept in Sweden and cars are built in the small town of Trollhättan, where it all started (Saab USA, 2006).
After more than 4 million produced vehicles, Saab cars are still influenced by aircraft design features. This can be seen and experienced in the cockpit-like ergonomics, the green illuminanced instruments and the need-to-know information displays. Another thing that has lived on from the time as an aircraft company is the great focus on safety (Saab USA, 2006).
1.1.2 Optis
Optis is the creator of the software Speos, which will be used during this project. The company was founded in France in 1989 and is since then supplying manufacturers with lighting system design. Optis offers software for light simulation and also engineering consultancy services. They have more than 1200 costumers, mostly in Europe, USA and Asia (Company history, 2008).
1.1.3 Problem Statement
The main purpose with this work to determine if the simulation software Speos can be used to simulate veiling glare in a realistic and correct way, to see if Saab would be able to use it for their needs. It was also important to access how easy or hard the software is to learn. To be able to validate the results from Speos, other data such as luminance and contrast ratio was collected by measuring the ghost images in the same light and weather conditions as simulated.
To understand the real problem with veiling glare, a driver study was carried. This was done both in real traffic (outdoor environment) and indoors in a simulated environment. The driver study investigated the importance of intensity and position of the veiling glare. A main interest was to come up with a veiling glare intensity level that people think is acceptable. This was done by letting test subjects drive and give their opinions about different intensity levels. By doing this, the intensity level could be used as a target value in simulations with virtual dashboard prototypes.
Another part of the thesis work was to determine if ghost images caused by veiling glare are more disturbing in specific parts of the windshield.
1.2 Scope
This work has been limited to only investigate the reflection of the dashboard into the windshield. Thus, we did not investigate reflections in the side windows nor direct glare from the sun or oncoming head lamps. Finally, we did not go very deep into the effects of veiling glare or external contrast reduction but focused on the ghost image effects on the windshield, both in terms of location and intensity.
3
2. Purpose and Research Questions
This thesis work investigated the possibilities to visually analyze the reflections of new designed dashboards into the windshield, by using virtual prototyping and simulation. This was done using the Speos software, developed by the French company Optis. The purpose with this is to avoid disturbing glare which will lower the total driving experience and give Saab bad will. Another goal was to come up with an acceptance level for veiling glare intensity and to analyze the importance of its position.
The goal is that Saab can use our results and the Speos software as tools for virtual prototyping of future dashboards and simulator-based studies of their reflection characteristics in order to minimize related problems in future products.
The questions that we will try to answer are: • How can Speos be used by Saab? • Can veiling glare be predicted? • How can veiling glare be measured? • How are people affected by veiling glare? • Can a level of accepted veiling glare be found?
4
3. Theoretical Frame of Reference
This chapter contains the background theory that concerns this thesis work and that has been studied to be able to understand and answer the research questions. Terms and expressions are explained to give an understanding for the following chapters. The first two parts in the chapter are about light and the human eye. After that follows a section that is called “Visual reduction while driving” which is mostly about the main topic, veiling glare. The last part is about Optis and their simulator-based design software Speos.
3.1 Light
Electromagnetic radiation may vary in strength and in wavelength. The change of wavelength is described in the electromagnetic spectrum, figure 1. The visible part of the spectrum is between the wavelength 480 nm and 770 nm. This part of the spectrum is referred as “light”. This visible part is bounded between the invisible ultraviolet and infrared regions. (Pedrotti & Pedrotti, 1996).
Figure 1- Electromagnetic spectrum (Britannica, 2008)
There are different ways to describe the strength of the light. This is categorized into two different categories: radiometry and photometry. Radiometry is the way to measure electromagnetic radiation (this can be done on the entire electromagnetic spectrum) while photometry only applies to the visible portion of the optical spectrum. (Pedrotti & Pedrotti, 1996)
The interesting component in this project is the photometry category of the light measurement. Photometry takes into consideration the human eye, while radiometry involves purely physical measurement. The eye responds different on different colors, for example yellow will seem to be much brighter then blue with the same radiant power during the day. Photometry is measured according to how the human eye detects it, not how a neutral detector would measure it. Of course people’s eyes are not identical and will see this information differently. This is why CIE (international commission on illumination, see chapter 5.1.5) has created a standard response which is represented in figure 2.
5
The right curve, at the green-yellow peak, shown in figure 2 is the luminous efficiency of the eye during daylight (photopic vision). When the eye is adapted for night (scotopic vision) the curve shifts towards the blue peak at 510 nm, represented as the left curve. This effect, when our sensitivity is changed between day and night, is called Purkinjes effect (Pedrotti & Pedrotti, 1996), (Starby, 1992).
Figure 2 – CIE sensitivity standard of photopic and scotopic vision (modified image) (Wikipedia, n.d.).
It is not only the wavelength that explains how a person will react on light during night. A human eye will also be much more sensitive to the intensity during the night-time conditions. In figure 3 the sensitivity levels is displayed for photopic and scotopic vision. (Starby, 1992)
Figure 3 – Sensitivity during photopic and scotopic vision (Starby, 1992)
Photopic Scotopic
6
3.1.1 Illuminance
Illuminance is a way to measure the strength of the light in an environment. It is a photometric method to measure how much a surface is lightened (the intensity of the incident light, per unit area), see figure 4. The unit for illuminance is lumen per m2 or the SI unit lux (lx). Lux is latin and means light that emits from a light source (Starby, 1992).
Illuminance was often called brightness, but this lead to confusion and was changed. (Pedrotti & Pedrotti, 1996).
3.1.2 Luminance
Basically all the things a person can see are luminance. Light cannot be seen until it hits a surface and reflects towards an eye. Luminance is often described as the reflection or emission from a flat, matt surface. This is a measurement of the luminous power that hits the eye when looking at the surface from a particular angle of view. Luminance is the density of luminous intensity in a given direction. It explains the quantity of light that falls within a given solid angle and passes through or is emitted from a particular area. The SI unit for luminance is candela per square meter (cd/m2). (Starby, 1992)
Figure 4 – Relationship between luminance and illuminance (Starby, 1992)
Luminance is defined by:
(1)
Where:
Lv is the luminance (cd/m2),
F is the luminous flux or luminous power (lm),
is the angle between the surface normal and the specified direction, A is the area of the surface (m2), and
is the solid angle (sr).
3.1.3 To Measure Luminance and Illuminance
There are two ways to obtain the luminance and illuminance values, either to calculate it or to measure it.
Luminance Illuminance
7 Calculating
To calculate the luminance of a surface is quite hard. To get a correct result, the surface has to be completely matt and this kind of material is hard to find (Starby, 1992).
The formula to calculate the luminance is: 2 / m cd E L π ρ⋅ = (2)
Where E is the illuminance and ρ is the reflection factor. Examples of reflection factors are presented in figure 5 (Starby, 1992).
Reflection factor
White paper 0,80
Light tree 0,45
White enamel 0,85
Light grey enamel 0,60
Dark grey enamel 0,15
Concrete 0,25
Clean aluminum 0,90
Figure 5 – Reflection factors (Starby, 1992)
The formula to calculate the illuminance is: lux A
E= φ
(3)
Where φ is the luminous flux, which defines as the intensity of a source with visible light. This quantity is measured in terms of the power emitted per unit solid angle from an isotropic radiator (theoretical point source that radiates equally in all directions in three-dimensional space).
Measuring
There are two common ways to measure light with a photometer. The first way is to measure illuminance in the unit lux (lx). There are plenty of different instruments to measure the illuminance. One instrument to use is the universal instrument that can measure both luminance and illuminance. Illuminance is measured by placing a small sensor towards the light that should be measured.
An example when using illuminance would be when the outdoor light is measured and the value 35000 lx is received on a sunny day with a small amount of clouds. Then when measuring indoors the value 100 lx is received. When comparing these values the following result is obtained:
350
(4)This calculation shows that it is 350 times brighter outdoors than indoors. The human eye will not notice that it is that much darker indoors because the eye adapts to the situation. The eye adapts to the dark room and the sensitivity for the light is much higher. This is why it is
8
important to be very careful when comparing indoor and outdoor results. More about adaption can be read in chapter 3.2.2 (Persson, Photac, Hagner, & AB, n.d.).
A way to compare the outdoor and indoor result is to use the daylight quote. If we divide the indoor illuminance with the outdoor illuminance; 100/35000=0,00285. The so called daylight quote is obtained, given in percent. The quote shows how much of the illuminance outdoors that remains indoors when you made the measurement. It is important to remember that the daylight quote will change in different places of the room. It is also important to remember that the clouds might change on the sky which will give different results when measuring illuminance outdoors (Persson, Photac, Hagner, & AB, n.d.).
To get a value of the luminance, the reflections from a surface is measured with a photometer. You cannot always trust that what you perceive is the reality. A luminance meter would get the same value for the grey fields below, but an illusion will make them look different. The grey color on the lighter background will look darker than on the black background, see figure 6 (Starby, 1992).
Figure 6 – Example of same luminance (Starby, 1992)
It is wise to document what is measured and in what circumstances. Below some examples are presented which are important to document when collecting values with a photometer.
• Which visual remarks were noticed? • Why did you take these measurements?
• What did you measure? Lighting in a room? Luminance on a display? Daylight in a landscape?
• How did you measure?
• Where, for an example in a car, did you do your measurements? This is answered with words and a simple sketch.
• Where there any light recommendations for the situation? How did your values compare to the given values?
• Own thoughts.
• Describe with words (and images) if something is uncertain.
By writing these things down it is easy to create knowledge of what is obtained (Persson, Photac, Hagner, & AB, n.d.).
Black body absorber
The strongest reflection on a windshield occurs when the background is black. This is because the reflections created on the windshield are brighter. When the background is brighter the reflections will blend into the background. So when measuring a reflection it has to be done towards a black background to be able to compare different measured values. The reason why it has to be completely black is that it will be constant in all different light
9
situations. If a grey surface would be used, the color would change in different light conditions and it would not be constant. A blackbody is a great absorber. (Pedrotti & Pedrotti, 1996).
Luminance Contrast Ratio
There are often given recommendations for the relationship between the luminance fields in environments where people work. In an office environment the relation 5:2:1 is used. This means that the working surface should be five times brighter than the surrounding environment and the table should be two times brighter than the environment. An example of this is a bright paper placed on a light colored table, see figure 7. It is often the color of the object that determines the luminance. The working object should not be too bright either, this will give a blinding effect. More than three times brighter between two surfaces next to each other, will blind the user (Starby, 1992).
Figure 7 – Contrast relationship between surfaces (Starby, 1992)
3.1.4 Contrast
Conditions for a person to detect objects are either that the object and the background have different colors, or that there are contrasts. Contrast means that the surfaces have luminance differences. A high luminance difference means that the contrast is high and also that it is easy for a person to detect the object. If an object has a low contrast it might not be detected at all, or if it has to do with long time work, the low contrast can lead to tiredness or headache (Nyman & Spångberg, 1996).
10
The luminance contrast can be expressed with the formula LC, where LC stands for luminance
contrast, LB for background luminance and LO for object luminance. Observe that the
luminance contrast LC has no unit (Nyman & Spångberg, 1996).
(5)
Dark objects on a light background will give a negative luminance contrast and bright details against a darker background will result in a positive luminance contrast (Nyman & Spångberg, 1996).
The eye is sensitive to different contrast ratios. Contrast ratio is another word for luminance contrast. In figure 9 relative contrast sensitivity is displayed according to CIE. The luminance is adjusted so that the contrast that is supposed to be seen is just noticeable (Starby, 1992).
Figure 9 – Relative contrast sensitivity according to CIE (Kelley, Jones, & Germer, 2008)
3.1.5 Reflection
Reflection is a common problem today for all kinds of working surfaces. If a user is exposed for reflections, the person will probably change position to avoid the reflections which will give bad sitting-positions and body ache.
There are relationships between the amount of light incident on a surface and the light that is reflected from the lightened surface. See 3.1.3 for an example of this relationship.
There are three different types of reflections: Specular, Haze and Lambertian. If the reflection looks like a mirror, then it is the Specular reflection. The reflection from a common copying paper is more diffuse and this type of reflection is called Lambertian. Another example of a Lambertian reflection could be a wall that is painted with a matt color. Many people think that so called diffused reflection is the same thing as Lambertian reflection, but a diffuser is a surface that takes light energy away from Specular direction and distributes it into many other directions. There is also a third kind of reflection that is a combination of Specular and Lambertian, which is called Haze reflection. To better understand these different kinds of reflections a simple experiment can be performed. If a laser beam illuminate a very large white card in a very dark room the general very soft illumination of the whole card is the Lambertian reflection. The sharp point in the middle is the Specular reflection and the soft ball around the Specular reflection is the Haze component. To measure these different reflections accurately is very hard. This is showed in figure 10. (Kelley, Jones, & Germer, 2008)
11
Figure 10 – Experiment to define different reflections (Kelley, Jones, & Germer, 2008)
One way to measure reflectance is to use the BRDF-method which is explained in section 5.1.1. The result of one of these experiments is displayed in figure 11 (Kelley, Jones, & Germer, 2008).
Figure 11 – BRDF experiment (Kelley, Jones, & Germer, 2008)
The three different components of reflections could be illustrated as figure 12.
Figure 12 – Three different reflection components (Kelley, Jones, & Germer, 2008)
Example of materials with Specular reflection is gold, silver and polished aluminum. Typical material examples of Lambertian reflections would be matt paper, textile and snow. Finally an example of materials with Haze reflections would be semi-glossy paper (Starby, 1992).
3.1.6
Refract the spee will inc would b3.2 T
The rol for tran Light e variable that the which f receptor send the Visual depends • • • •Refractio
ion means t ed of the lig crease speed be when lighThe Hum
l of the eye nslation via t enters the ey e hole in the e right amou focuses the rs called ro em to the br acuity is de s on four m Size Time Luminance Contraston
that the ligh ght is chang d when ente ht passes th Figure 13
man Eye
e is to conve the optic ne ye through e iris called unt of light light on the ods and con rain via theFigure efined as th major factors e ht change d ged. It will ering a thin hrough a win 3 – Refraction
e
ert light into erve (Light: the transpa the pupil, s passes thro e back wall nes which t optic nerve 14 - The hum he ability fo s which are: 12 direction wh slow down nner materia ndshield. (S n (Kelley, Jon o sensory s Human eye arent protec see figure 14 ough the eye of the eye translate the e (Light: Hu
man eye (Ligh
for the eye
hen it passe n when it en al, se figure Starby, 1992
nes, & Germ
ignals. The e). ctive cornea 4. The func e. The light containing e light ener uman eye, 2 ht: Human ey to distingui es materials nters a more e 13. An ex 2). er, 2008) se signals a a and then ction of the p t then reach the retina. rgy into ele 2007). ye, 2007) ish subtle d . The reaso e solid mate ample of re are sent to t passes thro pupil is to a hes the flexi The retina c ectrical sign details. Thi n is that erial and efraction the brain ough the adjust so ible lens contains nals and s ability
13
The size of an object affects the size of the image on the retina. It is not the physical size of the object that is important, but the visual angle to the object. This means that a small object in a close distance will create a large image on the retina and can therefore be seen clearly. The eye also needs time to adjust for good visual acuity. Just as a camera needs more exposure time in dim light, so does the eye. When the eye gets enough time to adjust, it can distinguish objects at very low luminance levels. The two other factors that affect visual acuity, luminance and contrast, is presented in chapter 3.1.2 and 3.1.4.
3.2.1 Glare
Glare occurs mainly in two ways; too much light (measured with an illuminance indicator) and if the luminance range is too large. Both are disturbing for the affected person, but only one impairs the vision (Glare, 2007).
Discomfort
Discomfort glare occurs when the iris has adjusted to a very dark environment and then is exposed for a much brighter environment. A common example is to come out when the eyes have adjusted to the indoor luminance levels. The iris will adjust rapidly with a little bit of discomfort, but with no visual impairment during the adjustment (Glare, 2007).
Disability
In the second kind of glare the eye has adjusted to the average luminance of the whole field of view. A bright point of light in the view will not affect the average luminance in the room too much, but the light source directly into the eyes will lead to discomfort and disability to see. This will make a person turn away or cover the eyes to protect them (Glare, 2007).
3.2.2 Adaptation
The eyes have the capability to adapt to different light situations. To adapt from a dark to a light environment will happen in a couple of seconds, but when adapting from light to dark could take up to an hour. A common problem is when entering a tunnel, where the eye has to adapt from a light environment to a dark environment really fast (Starby, 1992).
3.2.3 Phototropism
Phototropism is the name of a phenomenon where the eye is drawn to light objects. This is often used in commercials to show specific parts more than others.
This is one reason why reflections could be dangerous and irritating, because often the reflections will appear as brighter objects in the windshield and the eye will wander to these light areas (Hemphälä, 2007).
14
3.3 Visual Reduction While Driving
There are a lot of factors that contribute to a reduction of visual quality for the driver. Common for all of them is that they decrease the visual quality by contributing to a reduction of all the contrast values in the road scene. This will make it much harder for the driver to detect different objects while driving (Mefford, Flannagen, & Adachi, 2003).
The most obvious reduction factor is the windshield. Even if the windshield is new and clean, it will reduce the visual performance for the driver to some degree. The road scene looks different through the windshield than the way it looks like from outside the car (Mefford et al., 2003).
Another thing that affects the vision for the driver is dirt and scratches on the inside and outside of the windshield. These factors will decrease the drivers contrast while driving in daylight, but they will also strongly reduce visual performance at night driving because of scattered light from oncoming headlamps (Mefford et al., 2003).
The last and maybe least obvious factor is what is called veiling glare. This phenomenon occurs when light reflects on the car dashboard into the windshield which creates ghost images of the dashboard in front of the driver, which strongly reduces contrast of the road scene (see 3.3.1) (Mefford et al., 2003).
A research study has been made by the transportation research institute at the University of Michigan, USA by Mefford et al. (2003) to see how much these different factors contribute to the total veiling luminance. The tests were made in sunny conditions which gave a high but still common level of veiling glare. The amount of dirt on the windshield was also assumed as normal. The tests were made using eighteen vehicles with different sizes. The result can be seen in figure 15.
15
The result shows that dashboard reflectance, in the road scene in this report called veiling glare, is by far the biggest cause to reduction of contrast for the driver. In second comes the windshield including scratches.
3.3.1 Veiling Glare
As mentioned earlier the phenomenon when light reflects on the dashboard into the windshield, creating a reduction in contrast of the objects and the background road scene, is called veiling glare. This phenomenon will make it harder for the driver to detect different objects while driving. That is because veiling glare lowers all the contrasts in the road scene. This is especially evident when a light dashboard reflects into the windshield.
Another type of veiling glare is disturbing reflections in the windshield, which is called ghost images. This occurs for the exact same reason as contrast reducing veiling glare explained above. The difference is that the problem is not mainly the reduction in contrast, but the disturbing effect of the reflection in the windshield. The phenomenon occurs because luminance differences on the dashboard reflect differently into the windshield. Stronger contrasts give greater risks for annoyance for the driver, but it also depends on the position of the reflection. The reflections to the right in figure 16 is often called ghost images.
Figure 16 – (Left) Example of contrast lowering veiling glare. (Right) Example of ghost images.
Veiling glare will always exist, at least as long as cars have windshields. But there are ways to reduce the amount of veiling glare and to make it as little disturbing as possible.
16
The main two factors that you can manipulate to minimize veiling glare are the dashboard and the windshields design. The amount of light that reflects into the windshield strongly depends on the windshield angle. Studies have been made to determine its importance and this can be seen in figure 17. But the fact is that most of the car manufacturers already are using windshields with angles less than 60 degrees, so much more cannot be done with this factor (Schumann & Flannagan, 1997).
Figure 17 - Reflectance depending on windshield rake angle (Schumann & Flannagan, 1997)
Another opportunity is to further adjust the windshield design. When discussing this with our supervisor at Saab, he told us that the windshield already is a very expensive part of the car and also hard to manufacture as it is today. It is possible to add a cover on the windshield that will decrease the reflections. The main reason why this is not an option for Saab is because it is expensive and does not work very well. This means that the only way we will be able to affect veiling glare in this work, is to experiment with the dashboard design, including different materials and design objects and investigating their impact on reflection intensity and positioning.
The way the reflected images appear for the driver also depends on a fifth factor, which is the brightness of the background environment. A reflected image against a dark background will create a very disturbing image that strongly reduces contrast of the road scene for the driver. A bright background on the other hand, will make the reflected images very vague and transparent, or the driver might not see it at all. This is a factor that can not be controlled in real life, so we just have to accept it and focus at the factors we can affect.
The worst possible condition for veiling glare would be a very sunny day when the driver is driving towards a very dark background. This condition is not very common, because most often the sun also lit up the surroundings. But one case when it could appear is just before the driver is entering a tunnel. The reason why this case were chosen was because this is an extreme case and the idea with this report was to see the lowest level of acceptance a customer would approve when purchasing a car.
17
3.4 Methodology
There are many psychophysical methods that you can use for research purposes in this area. We present three of them in this paper. The methods have different advantages and disadvantages which have to be taken into consideration when making a decision of which one to choose.
3.4.1 Simulation-based design
The use of simulation as a tool in product design is growing and may be seen as a natural step after the appearance of CAD-based product models in 3D. The purpose of the simulation step is to investigate the functionality of the product from different perspectives. A less advanced simulation step is to visualize the product and thus make it possible to assess the visual features of the product. The use of Speos, as in our project, is a typical application of simulation-based design, where a CAD-based model of a dashboard is imported and supplemented with more design features (like dashboard material), which results in a virtual prototype possible to study under different conditions.
Speos studies and other “technical” simulations do not need a specific simulator but are carried out on standard computer platforms. A more specific application is where there is a need for specific hardware, a simulator, and maybe also possibilities to include an operator. Typical applications here are found in aviation, where human-in-the-loop simulators have been used for design purposes since decades in addition to the even more traditional use for training. Now this trend is also coming up in the automotive area and the expression Simulator-Based Design (SBD) has been coined (Alm, 2007).
The main benefit of using simulation-based design is its contribution in the form of project time reduction and product quality enhancement. This statement is especially relevant if virtual prototyping is included in the process but it is also possible in many cases to include hardware prototypes in the loop. However, the production time for a hardware prototype is always longer than for a corresponding virtual prototype and does, according to Alm (2008) not allow for many design iterations but as a final step in an iterative design process this could certainly be worthwhile.
As will be demonstrated later in this report we have used a combination of these methods in our study on windshield veiling glare related to dashboard design.
3.4.2 Methods for data collection
In experimental design studies, where simulations or real world applications are used the question of what, how and sometimes also when to measure is crucial. Behind this there always is the basic question why, since the purpose of every measure must be very clear. This is a fact since data analysis is time consuming and to measure everything to be on the safe side is nothing to recommend.
There are two basic measurement principles, objective and subjective measures. In the automotive area, both these alternatives are used and often in combination, so also in our study. Objective measures are often related to performance at any level (e.g., the entire system including the driver or for some technical sub-system) but also for other specific purposes like, for instance, environmental light conditions. Subjective measures, on the other hand, are based on human assessment (Alm, Simulation-based design, 2008).
18
In the following we review some methods for data collection which have been used or considered in our project.
Method of Adjustment
This is perhaps the easiest method to determine a threshold value. The idea is to let the subject control the intensity of the stimuli. That means that some sort of control device has to be provided. If for example an audio test is held, the subject could control the volume with the control device, instead of telling the test personal what he or she thinks so that they can change the volume. This makes the test much easier and faster. There are typically two ways of doing a sound test. One way is to start with a clear sound and let the subject lower the volume until he or she does not hear anything anymore. The other way is to start with no volume and let the subject detect when a sound is noticeable. The two tests should be done a few times, where the threshold is the average value of the tests. The biggest disadvantage with this method is that you need to have an easily manipulated control device to control the intensity. This could be hard in many cases. A corresponding technique could be used for visual image control with a variation of light intensity (Psychophysical methods, 2008). Staircase method
The staircase method has both some advantages and a few disadvantages compared to other commonly used methods. When you perform a test using the staircase method, you start at a specific intensity level and you ask the subject a question. If you are doing a hearing test, the question should be: Did you hear the beep? If the subject answers yes, you will lower the intensity on the next stimulus, and if the answers is no, you will raise the volume (Cornsweet, 1962).
According to Cornsweet (1962) there are four decisions the experimenter has to take. At what intensity the test should be started, how large the steps should be, when the test should be stopped and when the series should be modified. Cornsweet suggests some rules about how this should be done:
• The test should start at an intensity level that is not too far from where you think the threshold will end up. The step size should be selected so that the test subject does not respond the same answer more than maximum four times. As in all psychophysical methods, this method is most efficient when the stimulus steps are the size of the differential threshold. That means that the steps should be the smallest step that the test subject can detect.
19
Figure 18 - Staircase method (Cornsweet, 1962)
• It is a little bit harder to decide when to stop the test. Typically the curve of the test will look like the one in Figure 18. There will be big changes in the beginning and then it will converge towards one value. Of course the test is more reliable with long series, but that will make it very time consuming. A decision has to be made which factor that is the most important for the specific test. The easiest method to decide when to stop the test is to determine the number of stimuli responses in advance. Another and better way is to determine in advance the number of trials after reaching the threshold plateau. That is because the number of trials before reaching the plateau is strongly dependant on the starting intensity.
• There are cases where the step sizes of the staircase method should vary during the test. If you for example are testing the visual ability in a dark room, the steps should be large in the beginning and decrease because of the dark adaption. The same procedure can be used when starting the test at an intensity level that is much higher or lower than where the threshold will end up. Large steps can be used until getting close to the final level and after that, it is better to use small steps.
• The Staircase method is a very efficient method, because it requires very few stimuli to reach the threshold value, compared to other methods. Another advantage is that is quite easy to use, but that is also its biggest disadvantage. Because it is so easy, the tested subject could figure out how the procedure works, which can have impact on the test answers. There is another similar method that is not as easy to predict.
20 Double Staircase Method
With the double staircase method you use two curves simultaneously where one of them starts above and one below the threshold value. By using the two different curves randomly, it will be very hard for the test subject to understand the system. The advantage is that you still have a very efficient method and also test answers that are little affected by the biases of the subject (Cornsweet, 1962).
Figure 19 - Double staircase method (Cornsweet, 1962)
Questionnaire
When performing a study it is often important to let the participants be anonymous. To get a view of the participants of the study, a questionnaire could be used to receive information about the test subject.
A questionnaire could also be useful to find out about how the test subjects experienced the test. It is also a great tool to collect knowledge about the participant’s experiences within the area which the questionnaire handles.
Objective measures
During this project we have used a Hagner S3 to measure the illuminance of the weather conditions and the luminance to get a value of the intensity of a ghost image. To do this a light trap was placed in front of the windshield and the luminance of the ghost picture was measured with a photometer from the inside of the car.
21
4. Method
The chapter begins to explain the start of the thesis work, its content and how the realization was planned to be able to solve the problems and answer the research questions. It also contains the methods that we have used to perform the work.
4.1 Project Realization
This thesis work started in October, 2007. The order of realization was strongly influenced by the cooperation with the French company Optis. The reason for this was that Saab had bought consulting favors from Optis that lasted during 2007. That meant that a big part of this work had to be finished during the year of 2007. There was a license deadline for the Speos software at the end of the year, and a license deadline for the viewer in the middle of January. Because of that, there was little time to begin with a proper literature study as is usually done. To be able to validate the results from Speos, our plan was to take photos and do our own measurements of veiling glare in a real environment. The logic order would have been to photograph, do our measurements and then simulate the exact same case with the software. But the lack of time forced us to do it in the opposite order, with the disadvantage that it was difficult to do the real world measurements in the same weather conditions as in the simulation sessions.
A main focus was to come up with a target value, contrast acceptance, for Saab to use for future dashboard design. A driver study was planned to be done to let a number of drivers perform a test where they gave their opinions about what intensities of veiling reflections in the windshield that they could accept. A concern was whether the test should be done outside in the real traffic, or inside in a laboratory. Another thing that was of interest to be investigated in the driver study was the importance of position. By doing that it would be possible to appoint if there are positions in the windshield where reflections would be more disturbing than on other places of the windshield.
To summarize the plan to perform this work, we first of all wanted to see how easy or hard it was to learn and use the simulation software Speos. This was of course an important factor for Saab in the decision whether they would buy the software in the future. Another issue in interest was to test if Speos simulates the reality in a realistic and correct way. This was done by comparing the simulated images with photographs of the same case, and also by comparing the luminance values in Speos with our own measurements.
Saabs hopes is in the future to be able to simulate new designed dashboards made in CAD, to see what they will look like in reality, i.e. virtual prototyping. By doing so it should also be possible to control that the levels of reflections from the dashboard to the windshield, did not exceed their maximum luminance target value. The major benefit here should be that Saab could detect weaknesses in the design at early states. Therefore, one of our goals was to come up with a luminance contrast value which drivers experience as acceptable. That was the first part of the Driver study.
The Driver study also included a position test to validate whether there are areas on the dashboard where the designers have to be much more restrictive with shiny parts and materials. There might also be areas where they can place objects without disturbing the driver.
22
To get a structure of the problem, the project was divided into three parts. An overview of the project can be seen in figure 20.
Figure 20 - An overview figure of the realization of the thesis work
4.2 Startup
To be able to understand the underlying theory for the phenomenon of veiling glare, the thesis work started with learning about the topic. Information was searched in books, on the internet and by talking to experts in the specific areas.
Because of an early deadline with Optis, it was not possible for us to control the order of realization for the thesis work ourselves. Therefore the first session of reading and learning about the underlying theory of the thesis work, was quiet short. We quickly had to learn about basic optics to be able to perform the first part of the thesis work. During the cooperation with Optis and also afterwards, the theory studies that had been interrupted earlier continued. A visit to Saab in Trollhättan was done in the beginning of October to meet our supervisor and to get an introduction to the thesis work. Another important event during that visit was a meeting with the area sales manager and the sales vice-president from Optis, to get an agreement about how the cooperation between them and us would work out. Usually you need to participate in a two week training to be able to use the software by yourself.
RESULT/DISCUSSION
ANALYZE Analyzed and compared resultsPROBLEM SOLVING
SPEOS Learned about Speos Prepared for simulations Performed simulations MEASUREMENTS Made own measurements and documentations DRIVER STUDY Studied and learned how to make a driver study Performed pre study Performed driver study in car driving simulator Performed driver study in real trafficSTART/PROBLEM
INFORMATION Searched information PROBLEM/TASK Received/discussed problem Handled problem Made plan for how to solve problem23
Therefore they just wanted us to decide what we wanted to have simulated, and then they would make the simulations for us. That was according to us a bad solution, because a big part of the thesis work was to experience how difficult the software is to learn. Finally they were convinced that it was possible for us, with some consulting from them, to perform the simulations.
4.3 Staircase method
We chose this method because it was a reliable and quick method. The only disadvantage with the method was that the test subjects could easily understand the scheme of intensity changes, which could affect the results.
4.4 Questionnaire
A questionnaire was used to collect personal data about the drivers and their experiences of ghost images. The answers from the questionnaire were then analyzed to see relations between the drivers and the results. The driver’s answers were also used to find further interesting topics to investigate.
24
5. Apparatus
This chapter describes the light-simulating software Speos. It also contains other subjects that have to do with the simulator-based design that is handled in this project.
5.1 Speos
Speos is the name of Optis software that is used for virtual prototyping for lighting systems. Speos can be used as a standalone software, but is most efficient when it is integrated in the CAD softwares Catia V5 or OptisWorks. The software can be used in numerous ways. For example you can visualize what an operator, pilot or driver will perceive. It can be used to simulate lit or unlit appearance of a lighting system. You can also use it as an analysis tool for contrast, glare, reflections, visibility and obstruction.
Figure 21 – Screenshot from Speos (Optis World, 2008)
The software is used by companies from all over the world and in many industrial fields. They have customers in areas such as electronics, automobile, aerospace, defense, the luminaire industry and traditional optics. (Optis World, 2007)
5.1.1 Squale
Optis uses a tool called the Squale to read different materials and determine how they interact with light. Squale stands for surface quality extractor. Because Optis virtually want to simulate what objects look like in reality, they must know exactly how different materials interact with light. By reading the material in a way called BRDF, they know how the material reflects, absorbs and transmit light. The information is stored in a file that Speos use when simulating how the light bounces in the environment.
25
Squale is a tool that can be brought inside a car and make measurements of the different materials. It is placed against a material and with a press of a button it will read in the material in just a few seconds. The tool sends out light from the whole spectrum in different angles against the material and measures the direction and amount of the reflected light. The Squale is connected to a computer that stores all the information from the measurement in a file. This file contains what is called BRDF (Squale, 2008).
BRDF
BRDF is a function that describes what a material looks like for the human eye. It is a way to describe how light interacts with different materials. The shortening stands for; bidirectional reflectance diffusion function, and it is about the interaction between light and objects. (Rauwendaal, 2004).
The BRDF function depends on the direction of the incoming light and the outgoing light. It also depends on the amount of the incoming and outgoing light. The function describes how much light the material is reflecting, absorbing and transmitting. Different wavelengths reflect, absorb and transmit different depending on the physical properties of the material and the BRDF is therefore also depending on the wavelength and the material properties (Rauwendaal, 2004).
5.1.2 Ray Tracing
Ray tracing is used in 3D computer graphics to create realistic images, computer games and movie scenes. The technique can handle complicated optics like reflections and refractions and therefore creates high quality photorealistic environments and images. The idea is to mathematically visualize images by following the light beams backwards. That means from the eye through each pixel of the screen, taking in count the contribution of each light source in the environment to that pixel. If a light beam happens to intersect with an object in the scene, the pixel will update and the beam will either recast or die out depending on how many times it is programmed to bounce (Ray tracing, 2007).
There are different algorithms to mathematically solve the rendering equations, and each way has its benefits. One of these algorithms is called Photon mapping (Ray tracing, 2007).
5.1.3 Photon Mapping
Packets of light called photons are sent out from the light source instead of tracing the light from the eye. When a photon hits a target, information about incoming direction, intersection point and the energy of the photon is saved in what is called a photon map. The outgoing direction of the photon is decided by the surfaces BRDF (see BRDF under 5.1.1). The programmer decides when the photon should stop bouncing (Photon mapping, 2007).
5.1.4 Viewer
When analyzing a simulation, Optis offers a viewer program that comes with Speos, where the simulated images can be opened. The viewer has many possibilities to analyze the image depending on specific interests. With the viewer it is possible to get luminance values from each pixel in the image, or with a click on the mouse convert the photo realistic image into a photometric image.
26
5.1.5 CIE
The different skies that are used in Speos are based on definitions made by CIE. The commission international de la Eclairage or international commission on illumination is a non-profit organization founded in France in 1913. They provide information about light, lighting, color, vision and image handling (Daylighting, 2007).
Figure 23 - Examples of CIE sky models (Daylighting, 2007)
CIE have used mathematical models to describe skies, for example clear, overcast and uniform. A clear sky has a visible sun which of course gives a luminance distribution that is much brighter around the sun. An overcast sky is cloudy and the sun is not visible. The distribution of the luminance is symmetrical around the zenith and the radiation from the sun makes zenith the brightest spot decreasing towards the horizon. A uniform sky is, as the name sounds, a sky with a uniform luminance, see figure 23 (Daylighting, 2007).
5.2 Smart Eye
Smart eye is a Swedish company that was founded in 1999. The company develop eye tracking devices, mostly for the automotive industry. With the smart eye system, the purpose is to track head movement, eyelids and gazing of the driver (Company, 2008).
The system is used to analyze the driver in different ways. By tracking the head movement and the eye gazing, you can analyze things like how the driver reads the road scene and how often and during how long time he looks at the speedometer. This information can be used in many ways. By being able to track the distance between the eyelids, the system can be used to warn when the driver is about to fall asleep. The main key to be able to do this in an effective way, is the computerized analysis of video images (Technology, 2008).
27
Figure 24 - Camera and IR-flash illuminators (Technology, 2008)
The system uses one or two cameras depending on what you want to analyze. The system mainly uses its own illumination by using a number of IR-flash illuminators. These use a frequency that interferes as little as possible with the outdoor light. That makes the system reliable in every environment. A camera tracks a number of facial features and uses a 3D model of a face for matching the features. In this way the position of the head can measured with high accuracy. The camera also detects the irises and the pupils which together with the head position makes it possible to track the gazing. Finally it also tracks the distance between the eyelids by using a 3D model of the eyeball, see figure 25 (Technology, 2008).
Figure 25 - How the Smart Eye system works (Technology, 2008)
5.3 Driving Simulator
In 1996 the industrial ergonomics division (IAV) started a virtual reality and simulation laboratory at Linköping University. It was started to support design-oriented research and educational activities in the area of Human-Machine Interaction (HMI). Today the system is based on a PC platform and is focused on the vehicle area such as aircraft and ground transportation. This means that future in-vehicle systems (IVS) can be implemented in the simulator and validated in a complete environment, that is a complete vehicle system and a realistic traffic scenario (Simulator, 2007).
The simulator hardware consists of five screens and video projectors for an environment presentation that give a 200 degree field of view. There are two different cockpits, one Saab
9-3, and sound f can con by the c automo and eva systems 2007).
5.4 P
There a used a H is place displaye When m represen luminan When m value w an aver5.5 L
A light with a h will ent being re light tra which i car in f window d one Scan from the mo ntrol the sim company A otive, teleco aluation of s, lane depPhotome
are plenty o Hagner S3, ed in the en ed in lux. measuring nts a measu nce at the gi measuring l within the ci age value oLight tra
trap is a co hole, covere ter the box eflected ins ap is used is done in th front of the w towards th nia truck co otor. The sim mulations. T CE Simulat m, defense, f different parture wareter
of tools to m showed in nvironment luminance uring angle o iven surface Figure 26 – luminance ircle. It is th of the entireap
onstruction ed with a da x through thside the box when meas he driver stu e driver. Th he light trap ockpit. A so mulator is o The driving tion. ACE i , media etc systems. E rning, blind measure both figure 26. W where the with a Ha of one degr e (Persson, – Hagner S3 ( with the H herefore imp area (Perss used to cre ark diffuse he hole and x. This mak suring on a udy in this p he luminanc p. 28 ound system overviewed simulation is a consulti etera. The s Examples ar d spot dete h illuminan When meas light will b agner S3 th ree. Within Photac, Hag (Persson, Pho Hagner S3, t portant to ta on, Photac, eate a comp material on d the dark d
kes the surf a transparen project. The ce was mea m gives the from a cont software is ing compan simulator h re adaptive ction, cock
nce and lum suring illum be measured here is a b this angle th gner, & AB otac, Hagner, the value th ake a coupl Hagner, & letely black n the inside diffuse wal faces inside nt surface t e light trap asured from driver a re trol room w called ASim ny, active in as been wid e cruise co kpit concep minance, dur minance with d and the il black spot he instrume B, n.d.). & AB, n.d.) hat is meas le of values AB, n.d.). k surface. T . Only a sm lls will prev e the box co to get a co was placed m the driver ealistic feeli where the su m and is de n many indu dely used in ontrol, nigh pts etc. (Sim
ring this pro h this tool, lluminance in the mid ent is measu sured is an on a surfac This is simpl mall amount
vent the lig ompletely b nstant back d on the hoo r’s seat thro ing with upervisor eveloped ustries as n design ht vision mulator, oject we a sensor value is ddle that uring the average ce to get ly a box t of light ght from black. A kground, od of the ough the
