ETI Project

Abstract
Self-deriving cars are developed by various companies like Tesla, Audi,

and Google etc. The idea of autonomous cars are originated by the arising of
accident due to careless people driving and in this way harmful accident
extremely occur.
The purpose of our project is to add-on a safety feature on vehicles that
is lane detection so that vehicle may not distract from the road when driver is
not in his senses. This can make car driving robust and easy for anyone and a
lot safer than before. For this purpose we have gone through many articles and
read papers on it. Many companies and Developers are interestingly
participating their part in this field. What we did is that we have used modern
Image processing techniques in combination with Machine Learning models.
We are using Open-CV python to overcome our goals.
We started with a simple lane detection module based on Open-CV
python to get our edges and lines in the image detected but the most
important challenge was to interpret these lines with the image to make sense
of lanes. For this we have used slop-intercept methods. Then the race begin for
integrating our code to the hardware. And eventually it became unique and
innovative by introducing mathematical calculations. For calculations we have
used Numpy and Math libraries. Python-Math provides many useful built-in
methods for advance calculations like area under the curve and arc-tan curves.
And Arc tan proved to be more efficient that slope intercept calculations. Some
expertise with arduino helped me to tackle GPIO pins of Raspberry pi to
control movements of the car. We have tried to make the code as simple and
optimized as possible. Rest of the details are below in detail.
INTRODUCTION
Self-driving cars is a dream of developers since many years and many
companies and individuals are taking part in its improvements day by day. We
also decided to put some efforts according to our knowledge in this global
development project.
Self-driving cars are software based cars that runs own their own. Self-
driving cars are an advance technological development in the field of
automotive industry that provides both comfort and safety features for
drivers. Basic module of this vehicle are lane detection and object detection.
Many popular self-driving vehicles have features like they take control when
they observe miss mindedness sleepiness of drivers. According to a research by
ASIRT (Association for Safe international road travel) on average 3,700 people
lose their lives every day on the roads. An additional 20-50 million suffer non-
fatal injuries, often resulting in long-term disabilities (WHO, 2021). This mostly
happens due to human mistakes. To avoid such mistakes self-driving cars come
to the ground.
According to Gringer Bonnie’s debate on “History of the Autonomous
Car” In titlemax.com the first self-driving model that was proposed by General
motors in 1939 that was guided by radio-controlled electromagnetic fields
generated with magnetized metal spikes embedded in the road. In 1958 GM’s
design was produced on commercial level and car’s front was embedded by
coils that were used as sensors. In 1969 John McCarthy put forward his
thoughts on self-driving vehicle where he proposed an Idea of lane detection
using camera.
observations
self driving
Environment through
algorithms
cameras
Movements
Decesion of
According to
Actions
actions
Not every method calculates the angle, some follows the color approach
and some follows similar pixels approach. But we will be using angle approach.
With the invention of self-driving cars manual driving problems are being
resolved to a large extent. A simple example is that machines never goes to
sleep. But developing self-driving cars increases the problem of large &
complex calculations and feature extraction at high definition. To resolve such
problems modern libraries like math-python and the most popular image
processing library Open-CV python are not less than a blessing these days.
If we look into the past self-driving cars were equipped with complex
electromechanical parts and required a lot of requirement specifications, just
like the first ever self-driving model that was proposed by General motors in
1939 that was guided by radio-controlled electromagnetic fields generated
with magnetized metal spikes embedded in the road. To drive such a car you
will need to rebuild the roads that should contain magnetized metal spikes.
Then in 1969 John McCarthy is an inventor who invented the modern
autonomous vehicles that is based on the artificial intelligence technology.
Development problems
First problem I had faced during development is ‘installation of

open-cv python’ in raspberry pi. I tried almost every effort according
to my knowledge but it didn’t installed. Then I got a method having
20 steps to install that took me almost a full day and it worked.
Second problem is that raspberry pi operates on 5v but all
electronics components like motor driver and servo motors works on
12v to meet this problem I have used Arduino to power the parts
using impulse generated by raspberry pi.
To meet the problem of coast we have tried to develop this
service in our region using cheaper things like camera. This will
reduce the effect of customs and also of expensive sensors. Once this
technology gets into the market an automatic competition will grow
up. And market competition of large companies will automatically
introduce awareness. By development perspective we are using
camera and are aimed to do every-thing using programming
techniques rather than using expensive sensors and just
manipulating the values.
Objectives and scope
 To make daily driving safer

 To prevent car theft
 To save time by introducing features like auto parking.
 To convert normal vehicles to level 2 automated cars by
introducing a kit for every car.
 To put technology into market
Hardware Specification
 Raspberry Pi
 Arduino
 L298N driver
 Robot car chassis
 9V Batteries
 Battery connectors
 7” LCD
 Power bank 5V-3A output
 Jumper wires
Software Specifications
 Raspbian OS
 PyChram IDE
 Arduino IDE
 A windows PC / Android OS
 Python 3.7 no less or more
Libraries
 Numpy
 CV2
 Math
 Tensorflow
 RPi.GPIO
 Picamera
 Time
 Matplotlib
 Pillow
 grab-screen
 Collections
 IO-python
Functional Requirements Non-functional Requirement
 Should detects lanes and show them on Maintainable
the monitor
 Move according to lanes Vehicle should be Reliable
 In future Should add detects Obstacles Integrity

module
 Should be controlled wirelessly by any It should be reusable
device on the same network.
 Have functionality of dash-cam Legal Compatibility
Optimized Performance
ADVANTAGES:
1) The self-autonomous vehicles has reduced the number of
accident on the road.
2) Accident on the road the higher level of autonomous
vehicles are reduced
3) Dangerous driver behaviors.
4) Reduction in traffic deaths.
5) Consumer saving.
6) Reduction in travel time.
7) Improvement in fuel economy etc.
Disadvantages of Self-Driving Cars
1) Jobless Drivers
2) The price of the self-driving cars
3) The possibility of even worst crashes
4) No one is guilty
5) Privacy concerns
6) Security worries
7) Problems related to weather
Conclusion
In this project a simple prototype of a self-driving car is

demonstrated. The prototype car can recognize three separate
signals – turn left, turn right and stop with great accuracy
and take decision accordingly. The prototype can be further
improved by including more signals and using ML approach
References
 Abbas, A. N. (2021). Experimental Analysis of

Trajectory Control Using Computer Vision and Artificial
Intelligence for Autonomous Vehicles.
paperswithcode.com.
 Google
 Wikipedia
 Textbooks

ETI Project

Uploaded by

Copyright:

Available Formats

ETI Project

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ETI Project

Uploaded by

Copyright:

Available Formats

Abstract

Self-deriving cars are developed by various companies like Tesla, Audi,

First problem I had faced during development is ‘installation of

 To make daily driving safer

 Move according to lanes Vehicle should be Reliable

 In future Should add detects Obstacles Integrity

In this project a simple prototype of a self-driving car is

 Abbas, A. N. (2021). Experimental Analysis of

You might also like