A

Seminar Report

on

“PROGRAMMABLE AUTOMOTIVE HEADLIGHTS"

Submitted by,

Patil Rushikesh Suresh

Bachelor of Mechanical Engineering

Guided by

Mr. V.V.Gundge

DR.BABASAHEB AMBEDKAR TECHNOLOGICAL UNIVERSITY,

LONERE

Department of Mechanical Engg.

PES’S COLLEGE OF ENGINEERING PHALTAN,

2021 - 22
 CERTIFICATE
This is to certify that Mr. Patil Rushikesh Suresh has successfully completed the
Seminar work entitled “Programmable Automotive Headlights” under my

supervision, in the partial fulfillment of Bachelor of Engineering -

Mechanical Engineering, by Dr.Basaheb Ambedkar Technological University Lonere.

Prof. V.V.Gundge. Prof. G. V. Thombare . Dr. M.A. Natu

Seminar Guide. HOD. Principal

Seal/Stamp of college Principal

Place: Phaltan education society’s college of engineering,

Phaltan
Date:
 ACKNOWLEDGEMENT

I wish to express sincere thanks and gratitude to the Mr. G. V.

Thombare (H.O.D) and Mr.V.VGundge sir(Seminar coordinator) for the whole
hearted encouragement, advice and Timely guidance in completing this seminar
work.
I also like to thank my family for supporting me through this seminar activity. It is
my pleasure to thank Mr. M. A. Natu (Principal College of Engineering, Phaltan) for
giving me this opportunity.

Rushikesh Suresh Patil

(1967661612036)
B.Tech mechanical
 CONTENTS

Sr. No. Description Page

1. INTRODUCTION 1

1.1 Reasons For Use of A Projector 2

2. LITERATURE SURVEY 3

3. OBJECTIVES 4

4. METHODOLOGY 5

4.1 Stages of Processing 5

4.1.1 Object Analysis 5

4.1.2 Shadow Placement 5

4.2 Algorithm 6

4.3 Design of Two Tracking Algorithms Using Sensor Array 6

4.4 Development of Intelligent Front Lights 7

4.5 Lane Recognition and Track 8

4.5.1 Region segmentation based method 8

4.5.2 Feature Based Method 10

4.5.3 Model Based Method 11

4.6 Vehicle Recognition (Light Recognition) and Track 11

4.6.1 Feature Based Method 12

5. EXPERIMENTAL STUDIES 13

6. ADVANTAGES 14

7. DISADVANTAGES 14

8. RESULT AND DISCUSSION 15

9. FUTURE WORK 16

10. CONCLUSION 17

i
11. REFERENCES 18

ii
 LIST OF FIGURES

Figure Illustration Page

1.1 Anti-glare headlights / high-beams 1

4.1.1 Stages of Processing 5

4.3.1 Flow Chart for Tracking of Vehicle 7

4.5.1.1 (i) Non-evenly divide the interested region 8

4.5.1.1 (ii) Binarize every window 9

4.5.1.1 (iii) Pick out interested windows that include lane 9

markings

4.5.1.1 (iv) The final lane markings generated 10

4.5.2.1 Feature based lane marking detection 10

4.5.3.1 Model Based Method 11

4.6.1 Front and rear lights of a vehicle 12

4.6.1.1 (a) Colour, size, shape, symmetry, and etc. feature 12

based recognition

5.1 This figure is divided into four main regions 13

iii
 NOMENCLATUR E

Abbrevations
• EKF - Extendedkalman Filter
• UKF - Unscented Kalman Filter
• HID – High Density Discharge
• AFS – Adaptive Front-lightning System
• LDR – Light Dependent Resistor

iv
 ABSTRACT

The anti-glare headlight enables vehicle drivers to take full advantage of their high
beams without the fear of blinding oncoming drivers with the glare that occurs
while driving. The primary goal of an automotive headlight is to improve safety
in low light and poor weather conditions. But, despite decades of innovation on
light sources, more than half of accidents occur at night even with less traffic on
the road. Recent developments in adaptive lighting have addressed some
limitations of standard headlights, however, they have limited flexibility -
switching between high and low beams, turning off beams toward the opposing
lane, or rotating the beam as the vehicle turns - and are not designed for all driving
environments.

v
vi
 INTRODUCTION

Headlights play a crucial role in safe driving. The National Safety Council
reports that three times as many road fatalities happen at night-time because of the
difficult driving conditions. Headlights should emit a steady beam of light and be
cleaned regularly for optimal performance. Glare from headlights, especially high
beams of oncoming vehicles cause significant stress and distraction at best and
temporary blindness at worst. Trucks and other vehicles with headlights at high
positions are the worst offenders. Although glare is not often reported as a cause of
accidents, hundreds of fatal night crashes attribute glare as a contributing factor every
year. Glare is especially problematic for the elderly whom take eight times longer to
recover from glare as compared to a 16-year old. Although high beams are a nuisance
to other drivers, they are beneficial on narrow, curvy, and poorly lit roads, especially
in rural areas where wildlife routinely approaches the road.
More than 300,000 crashes and thousands of fatalities are caused by rain and
snow at night annually. Approximately 30% of drivers are stressed by glare causing
hundreds of fatalities every year. Thus, a headlight that adapts to the environment can
be critical to improving safety on the road during poor visibility conditions.
Traditional headlights consist of a small number of lamps with simple optics to direct
a light beam onto the road. Starting with gas/oil lamps in the 1880s, research has been
primarily geared towards developing headlights that can be electrically controlled,
have a long working life, and are bright and energy efficient.

Fig 1.1: - Anti-glare headlights / high-beams

 1
1.1 Reasons for Use of a Projector

a) Accurate Light Dispersal :

The elliptical reflectors in the projector headlight housing design is a
powerful way to get a focused beam and cut down on "scatter loss" on the road.
b) Style Accommodation :
Those who are advocates of the projector headlight point out that using
this kind of headlight can enhance the look of the vehicle's front end and give
it a more aesthetic appeal.
c) Lens Based Car Lighting :
As the traditional parabolic reflector style of headlight began to give
ground to newer lighting technologies, projector headlight design became part
of the set of headlamp options that make use of lens positioning to provide
great setups for night driving.
d) Division of Light Beam :
The projector headlights are divided into smaller units of light, which
is at utmost requirement for our system.
 2

2. LITERATURE REVIEW
1. Kalman Filter-Based Tracking of Moving Objects Using Linear Ultrasonic Sensor Array
For Road Vehicles (2018) :
In this paper by, Shengbo Eben Li, Guofa Li, Jiaying Yu, Chang Liu, Bo
Cheng, Jianqiang Wang, Keqiang Li, presents a cost-effective approach to track
moving objects around vehicles using linearly arrayed ultrasonic sensors. And to
understand the detection characteristics of a single sensor, an empirical detection
model was developed considering the shapes and surface materials of various detected
objects.

2. Programmable Automotive Headlights :

This article by, Robert Tamburo, Eriko Nurvitadhi, Abhishek Chugh, Mei
Chen, Anthony Rowe, Takeo Kanade, Srinivasa G. Narasimhan, introduces an
ultralow latency reactive visual system that can sense, react, and adapt quickly to any
environment while moving at highway speeds. Our single hardware design can be
programmed to perform a variety of tasks.

3. Development of A Headlight Glares Simulator For A Driving Simulator (2013):

Here, Alex D. Hwang, Eli Peli, described the design and construction of a
headlight glare simulator to be used with a driving simulator. The system combines a
modified programmable off-the-shelf LED display board and a beamsplitter so that
the LED lights, representing the headlights of oncoming cars.

4. Anti-glare Headlight System :

In their paper, Chitra Balasubramanian and Lyle Ethan Mark Rodericks ,
explained that the anti-glare headlight enables vehicle drivers to take full advantage
of their high beams without the fear of blinding oncoming drivers with the glare that
occurs while driving.
Page No. 3
5.Automatich Headlight Beam Controller (2017):

In their research, Aslam Musthafa R, Bala Krishnan T, Seetha Raman N,

Shankar M, Asst.prof Ms. Swathi R, developed an automatic headlight dimmer which
uses a Light Dependent Resistor(LDR) sensor to dim the headlight of our vehicles to
avoid human eye effects. This automatically switched the high beam into low beam,
therefore reducing the glare effect by sensing the light intensity value of approaching
vehicle.

OBJECTIVES

• One of the best objectives of programmable automotive headlights is to prevent the

road accidents.
• And the other reason is to improve safety in low light and poor weather conditions.
• Adaptive headlights.
• Reactive visual system.
• Computational illumination.
• To eliminate the standard headlight system.

Page No. 5
4. METHODOLOGY
4.1 Stages of Processing

Fig 4.1.1: - Stages of Processing

This figure shows a simple representation of a system flow block diagram which
has three stages.

a) Capture the image and perform some pre-processing.

b) Process the image using the below processing block diagram.

c) Project the final output on the windshield of the opponent incoming vehicle.

4.1.1 Object Analysis:

Here we indentify a pair of headlights belonging to the same vehicle by image
processing. Headlights are detected using the assumption that they are the brightest
objects in the system’s field of view. A very short exposure (150-200 µs) time is used
and the image is threshold. We identify the high intensity regions of a given image
and find the coordinates of their centres. If the objects have similar Y- coordinates,
horizontal angle of less than 15° and equal intensity, it is considered as a headlight
pair from the same vehicle.
4.1.2 Shadow Placement:
The various stages involved in casting a shadow are,
• After the headlight pairs are identified, each of their centroids are parsed to this
stage.
• The distance between these cancroids’ is measured.
• Half the distance is calculated and taken as the side length of a square which is
plotted above the left headlight.
• In the Indian driving system the driver is seated at the right side of the vehicle.

Page No. 6
 So we now cast a shadow to the leftmost half of the incoming vehicle’s windscreen.
Casting a shadow is done by programming the projectors.

4.2 Algorithm
• Step 1: Get input from the camera
• Step 2: Increase the contrast of the image captured to infinity so bright areas become
brighter and dark areas become darker.
• Step 3: The image is threshold so areas of very high intensity become white and areas
of lower intensity become black.
• Step 4: Extract the white areas and form objects.
• Step 5 Retrieve 2 objects (A and B).
• Step 6: Compare object A and B and check if they are headlight pairs. If yes, go to step
8, else step 7.
• Step 7: Collect next object B if possible. If not, collect next object A if possible. If all
objects are exhausted and compared with one another, go to step 1.
• Step 8: If the two objects belong to the same pair of headlights then calculate the mask
position and size by the following calculations and project the mask.

4.3 Design of Two Tracking Algorithms Using Sensor Array

The summarized structure of Tracking algorithms’ flow chart is shown in
below fig. Single sensor modeling considering shapes and surface materials is marked
as black lines. Measurement error model, detection scope model and detection
probability model were built based on the work presented. The proposed Extended
Kalman filter (EKF) and an Unscented Kalman filter (UKF), tracking Algorithms
were presented.

Page No. 7
 Fig 4.3.1: Flow Chart for Tracking of Vehicle

4.4 Development of Intelligent Front Lights

The vehicle lamp has been developed from the original incandescent lamp to
halogen lamp and HID (high intensity discharge lamp) during the 20th century.
Nowadays, halogen lamp and high intensity discharge lamp have been widely utilized
in all kinds of vehicles. Meanwhile, as the LED (Light-Emitting Diode) vehicle lamp
matures, it is increasingly being used. Because LED lamp has energy saving,
flexibility, high efficiency characteristics, choosing it as vehicle front-light lamp is a
trend. At the same time, the intelligent lighting system has been changed from static
illumination to AFS (Adaptive Front-lighting System) and up to various kinds of
intelligent lighting control systems.
The AFS adjusts the movement of traditional front lamps both in horizontal
and vertical levels. The illuminated areas in front could be changed automatically
according to the vehicle conditions. The lamps adaptively adjust the horizontal
illumination area to the left/right front area of the roads while turning left/right and

Page No. 8
revise the vertical lighting level while the vehicle is in ramp road. In general, the
traditional AFS has experienced three generations, the static curved road illumination,
the combination of following lighting and static curved road illumination, and the
multi-function illumination which includes the basic mode, curve road mode, highway
mode, town mode, village mode and rainy day mode.

4.5 Lane Recognition and Track

Based on the current technology level, the reliability of the vision detecting
system in vehicle is mainly influenced by the factors such as shadows, weather, and
intensity of the illumination. Specially, in night situations, the valid image is more
difficult to fetch due to the front and rear lights of vehicles, reflections of constructions
and wet roads, etc. The algorithm for the night-time implementation would be much
more complex and very challenging. Basically, according to different recognition
algorithms, vision based lane recognition methods can be divided into region
segmentation, feature, model, 3D vision, multi sensor merging based methods, etc.

4.5.1 Region segmentation based method

Usually, the region segmentation based method evenly or non-evenly divides

the interested region into several small rectangular windows as showed in figure (i)
and then binaries every window as showed in figure (ii).

Fig 4.5.1.1 (i): Non-evenly divide the interested region

Page No. 9
 Fig 4.5.1.1 (ii): Binarize every window
The interested windows that include lane information can be picked out as
showed in figure (iii) according to the contrast information. After that, the lane
markings in each interested window can be linked together to generate the stepped
lines. Thus the lane characteristics in the interested area of image can be described by
these stepped lines so as to form the final lane markings in figure (iv). This method
could describe every kind of lane markings including straight lines and curves and has
good flexibility. However, some artificial noises might be added during the region
segmenting.

Fig 4.5.1.1 (iii): Pick out interested windows that include lane markings

Page No. 10
 Fig 4.5.1.1 (iv): The final lane markings generated

4.5.2 Feature Based Method

The feature based method may detect the lane markings by the start points of
lane boundaries, the forward trend, the gray values, etc. Based on this information, the
lane can be presented as some possible lane vectors. In these vectors, the vector that
has the lowest distance from the former vector can be selected to best express the lane
trend, so as to confirm the lane boundary. This mean has lower calculation time and
good real time performance, but can only be adopted in the optimal lane marking areas
as it often fails when the interferences are bigger.

Fig 4.5.2.1: Feature based lane marking detection

Page No. 11
4.5.3 Model Based Method
The model based method cites parabola (see figure v), hyperbola (see figure
vi), B-snake (see figure vii), even triangle model (see figure viii) to express different
lanes. After detecting the basic edge of lanes, a model curve with to-be-determined
parameters is provided to express the actual lanes. The parameters of this curve are
adjusted to optimally match the lanes. The curve with the best parameters is the one
that is picked out to describe the lane. But if the number of control points of the curve
is big, the processing time might be quite long.

Fig v: Parabola model Fig vi: Hyperbola model

Fig vii: B-snake model Fig viii: Triangle model Fig

4.5.3.1: Model Based Method

4.6 Vehicle Recognition (Light Recognition) and Track

In night situations, because of the severe dark conditions, vehicle recognitions
basically rely on the front and rear lights recognitions as showed in figure 21. Similarly
to the recognition of lanes, the vision based vehicle lights detection methods can be
divided into feature, lane detection merging, motion based methods, etc.

Page No. 12
 Fig 4.6.1: Front and rear lights of a vehicle

4.6.1 Feature Based Method

As showed in figure (a), feature based methods usually recognize the front and
rear lights by their colour, size, shape, symmetry, aspect ratio, etc. Usually, the front
and rear lights can be distinguished by their different colours. There are also some
constraints to identify the vehicle lights. For example, the vehicle lights to be
confirmed should be surrounded by a rectangle. The two white spots should be
symmetrical and in a reasonable region. Specially, the rear lights areas could be
constrained by the red light emitted from them, which can enhance the reliability of
rear lights detections. These different constraints could be considered as ideal
conditions to detect vehicle lamps.

Fig 4.6.1.1 (a): Colour, size, shape, symmetry, and etc. feature based recognition

5. EXPERIMENTAL STUDIES
We have simulated the experiment on MATLAB and have arrived at good
results. We’ve tested the simulation it in two ways. The sizeable number of drivers

Page No. 13
(about 30%) who experienced night time glare as ‘disturbing’ cannot be ignored. The
response data of drivers falling into this category was further analysed to bring out the
differences that might exist due to the type of glare and age and gender of the
respondent. In fact, the frequency distributions of the glare-disturbed respondents over
age and gender groups, for oncoming and following Glare, did not indicate any
difference due to the type of glare; the distributions were found in close proximity
with each other. The tested video has been shown in below fig.

Fig 5.1: -This figure is divided into four main regions.

a) A raw footage taken.

b) Detecting the light sources and finding their respective x-axis and y-axis.
c) Real-time shadow placement-is done by programming the adaptive headlights to
cast a shadow on the windshield of the opponent car.
d) The view of output from the headlight.

6. ADVANTAGES
• It’s a cost effective solution for the problem of night time accidents occurring
during two way roads.
• It can easily be mounted on the running cars and low cost cars as a necessary.
• Increased safety for drivers & pedestrians.

Page No. 14
7. DISADVANTAGES
• Continuous power supply from the battery for the sensory circuit.
• The light from street light, tail lamp of vehicle also fall on LDR and cause the
switching circuit to continuously switch to high beam and low beam.
• The set value selection is a complex one.

8. RESULT AND DISCUSSION

• Generally speaking, our programmable headlight is a low-latency reactive visual
system with many uses outside of the automotive field.
• It has the exibility of illuminating or disilluminating any fast moving object.
• An interesting application is studying the trajectory of fast moving objects or fast
events.
• Typically, to capture these types of images, an expensive camera is needed and the
room needs to be brightly lit causing a decrease in contrast. Instead, with our
system, only the objects of interest need to be illuminated.

Page No. 15
9 FUTURE WORK

completely switched on or off. It should be capable of adapting to the environment

to improve safety in poor visibility conditions.

• Moreover, the design for adaptive headlights should not be limited to a single

task. It should be capable of performing many different tasks to help the driver in
multiple road environments.
• Our headlight design provides unprecedented light beam control over space and
time. We have demonstrated the flexibility of the headlight for numerous tasks:
allowing drivers to use high beams without glaring any other driver on the road
and allowing better illumination of road lanes, sidewalks and dividers.
• Our prototype can quickly react to the road environment within 2 to milliseconds,
and, thus does not create any flicker to be seen by the human eye.
• Further research and development is needed to make the prototype compact to fit
within actual vehicle headlight compartments.
• Further engineering is required to make the system reliable in the presence of
vehicular vibrations and heat.
• Lastly, more sophisticated algorithms and reliable software need to be developed
before deploying our headlight design.

Page No. 16
10. CONCLUSION

• A comprehensive survey of vehicle intelligent headlight system development is

presented.
• Many factors such as road geometry, road conditions (intense light, wet, fog) or
glare from other light sources (vehicles, road and structure reflected lights, lights
along the road) may cause reduced illumination range or uncomfortable feelings
for the drivers in vehicle and other vehicles, which may lead to traffic accidents.
• An optimal vehicle intelligent headlight system should take as many parameters
from the vehicle and the surround environment as possible into consideration and
be able to adapt on many kinds of driving conditions to provide a maximum
illumination area of the street for the driver, without glaring other road
participants.
• By investigating the development of vehicle intelligent front lights, discussing
some latest key technologies and algorithms of lane detection and vehicle light
detection and providing some possible proposals we come to the conclusion that
the Matrix-LED system could make the system more flexible and more effective
and vision based vehicle intelligent front light system can improve the driving
environment at night.
• However, this kind of system is still a challenging task for many car makers and
auto parts manufacturers.

Page No. 17
 REFERENCES

[1] Alex D. Hwang , Eli Peli. ‘Development of a headlight glare simulator for a driving
Simulator.’ Transportation Research Part C 32 (2013) 129–143
[2] Aslam Musthafa R, Bala Krishnan T, Seetha Raman N, Shankar M, Asst.prof Ms.
Swathi R., ‘Automatic Headlight Beam Controller.’ International Journal of Trend in
Research and Development (IJTRD), ISSN: 2394-9333

[3] Chitra Balasubramanian, Lyle Ethan Mark Rodericks. ‘Anti-glare Headlight System.’

[4] Mr. Sandip S. Jadhav, Prof. Ansar A. Mulla, ‘A Multi Featured Automatic Head Light
Systems Prototype for Automotive Safety’, International Journal of Engineering
Research and Technology. ISSN 0974-3154 Volume 10, Number 1 (2017)

[5] Prof. Pratik Ashok Patil, Shubham Sunil Badave, Shubham Sunil Ingwale,
‘Automatic Headlamp Illumination Control System’, IJIRSET, Vol. 6, Issue 3,
March 2017

[6] Robert Tamburo, Eriko Nurvitadhi, Abhishek Chugh, Mei Chen,

Anthony Rowe, Takeo Kanade, Srinivasa G. Narasimhan., ‘Programmable
Automotive Headlights.’

[7] Shengbo Eben Li, Guofa Li, Jiaying Yu, Chang Liu, Bo Cheng, Jianqiang Wanga,
Keqiang Li. ‘Kalman filter-based tracking of moving objects using linear ultrasonic
sensor array for road vehicles.’, Mechanical Systems and Signal Processing 98
(2018) 173–189

[8] Vithalkar Akshay Ganesh, Khavare Vinayak Vithal, Maitshaphrang Syiemlieh,

Gawande Prashik Babarao, Supriya Y. Sawant, ‘Automatic Headlight Beam Control
System, International Journal Of Mechanical And Production Engineering’, ISSN:
2320-2092, Volume- 3, Issue-7, July-2015

Page No. 18