International Journal of Advance Science and Technology

Vol. 29, No. 10S, (2020), pp. 6979-6978

Road Sign Recognition and Vehicle Control Using Raspberrypi

Mrs.G.Annapoorani1, Yasam.Dharani2, C.Silpa3, P.V.Harsha4
1
Asst.Proffessor, SRMIST ,
2,3,4
ECE Dept, SRMIST

Abstract
This paper appearance into the advance of road sign recognition and vehicle management by
employing a Raspberry PI, it needs a computationally simple but accurate algorithm. It instantly
assists the automatic driving system in recognizing and police work traffic signs more effectively.
During this planned work, the system instantly acknowledges the road signs and controls the car by
giving commands in certain actions. In this system, we have got used the raspberry pi 3 processor
and web camera which mechanically captures the video around. First, the video information convert
into several frames that area unit processed by Sobel filter exploitation edge detection. Next, the
photographs of the road signs are converted into gray scale images, We present a concise overview of
image processing methods and analysis get and filtered in it by using Simplified Gabor
Wavelets(SGW), wherever the parameters are optimized and when we tend to extracted the Region of
Interest(ROI) using the foremost stable external regions area unit notice by using the algorithmic
program and classified the taxonomic category of traffic signs. The Survey results from German and
Chinese traffic sign databases show that the proposed methodology gets a comparable performance
with the state of the art method and even more in the future, the complete process of detection and
classification were updated as per the real-time operation potency. We tend to use Sobel filters for
detecting road signs and therefore the detail of a sensory system may be explained.
Index Terms: SGW, ROI, SOBEL FILTER, EDGE DETECTION

1. INTRODUCTION
Road Signs will provide one of the most vital records that will help for riding protection measures to
the drivers and different users. These street symptoms are classified into colors and shapes for
effortless identification. Many of the accidents are being used for lowering their attention of drivers
and they usually pick out to pass the street signs. When the climate stipulations have to like rain and
occasionally the turnings of area points can cause accidents mostly. So, the attention of the symptoms
has to be a tremendous help to minimize accidents and deaths. The enforcement of street signal
detection and recognition systems is the use of a photograph processing technology that will supply
make sure for each driver is being aware of regulations and hazards on the street and it will be
beneficial to decrease the variety of accidents and deaths. So to make it effortless for drivers to study
and recognize, these visitor signs are being designed with the aid of a precise structure of it and
colorings can acts act as symbols, so there is a small difference between the traffic signs and
background. For example, the speed restriction is 60 in visitor’s signs. However, traffic signs and
symptoms are no longer easy to follow, so there are many facets such as awful weather, perspective
variation, bodily damage, etc., are the difficulties that we are confronted with and overcome with this
system.
A vast vary of sensor technological know-how can be used such as Logitech cameras, an ultrasonic
sensor that can be used as a transducer that can be capable to sends and receives ultrasonic signals that
relay back to the facts like an object’s limit. For detecting the avenue signs and can ship the next path
to the raspberry pi which is related between Logitech digital camera and servomotors.
Though, the proposed system can also be designed by using Raspberry PI which can be using through
a low profile monitor and to use it as windows based systems using Linux simulation software. This
Logitech camera can work with a Sobel operator which is called Sobel-Feldman operator or Sobel-
filter, which is used for image processing and computer vision which can be collecting particular
high-frequency variations of an image. In this, for transferring the directions we use servomotors with
L239D motor drivers which can lets in DC motor to the driver in either direction. Servo motors are
rotatory actuators or linear actuators which will allow the precise control of angular and linear
positions, pace and acceleration, which will consist of an appropriate motor coupled a sensor for

ISSN: 2005-4238 IJAST

Copyright ⓒ 2020 SERSC 6979
 International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp. 6979-6986

position feedback. Our method can be achieved only the function extraction of the detection and
classification stage.

2. LITERATURE SURVEY
This paper describes the usage of automation motors in industries for transporting goods, loading,
and unloading specifically in the manufactureand exporting units. The important difficulty is to
introduce automation in nearly all fields of industry. This paper is to suggest a driverless battery car
which runs robotically barring assist of the driver and it is normally recognized as Automated Guided
Car (AGC), This courses us to selected spot automatically on its own by using sensing environment,
navigating around the area and fulfill the human needs, it can additionally explain the fundamentals of
operations of each unit whilst crossing that unique unit as it contains preloaded information.[2]

DRAWBACKS:
 Slight hard to handle as it needs a smooth and ramped area to move
This paper describes Driverless automobiles become current technologically advanced within the
transport trade. So, they are anticipated to transfigure the transport by crediting safer, a lot of
convenience, and nice potency. However, there are a lot of technical troubles with the region of
driverless vehicles, and disastrous outcomes are the protection issues that having public. Having
suited protection estimates within the region for driverless vehicles constitutes trouble to the accepts
the society, and one among the best prohibiting issue. This paper discusses regarding the self-
sustaining science of driverless motors and therefore the protection problems with this technology.
Exploring the present communication applied sciences of vehicular ad hoc networking, and the way
unreal neural networks and therefore the software-defined network will assist to protect the driver-less
automobiles from malicious attacks.[5]

DRAWBACKS:
 It delays while responding time
 It suddenly uses a break when the object is closed without any prior indication of the
obstacle
This paper describes driverless motors using a machine mastering algorithmic rule for deciding in
driverless navigation is developed. This approach is enforced by suggests that of adapting gift verified
convolutional neural networks to the cutting-edge drawback. Compared to the normal methods
supported by algorithms from pc graphics in camera work, this switch mastering technique tends to
optimize all process steps at the same time, consequently resulting to the higher performance whereas
conserving a excessive performance-to-cost magnitude relation.[6]
DRAWBACKS:
 The effective range and the factor lead time is very less
This paper may be a period unnoticeable coronary heart worth watching system was planned and
enforced. This device goals to monitor the heart rate of the passengers by using alower-priced digicam
inserted inside the car, which can be without delay embedded in the car’s rearview mirror. Also, they
are up to date this device with the main gadget of a check driverless automotive, and they suggest
however driverless cars got reply in response to serious clinical emergencies. Moreover, they review
however that system will take a bonus from the auspicious so ptions of Google I/O and Google AI.
The strategy relies all on Remote Photoplethysmography (RPPG), within which the coronary heart
worth is detected from the fragile small modifications going down within the pores and skin
coloration of the face for the period of every and each pulsation. The face is automatically detected
and half-tracked, then the raw signal is calculated from each frame over each 10-seconds in a sliding
window[9-11]. After that, a series of signal processing ways area unit applied to the raw alerts and
recovers the center worth frequency.[7]

DRAWBACKS:
 Decreasing the window length of the car degrades the accuracy
In this paper, we are visiting to study the consequences and network parameters and their

ISSN: 2005-4238 IJAST

Copyright ⓒ 2020 SERSC 6980
 International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp. 6979-6986

calculations in a position environment. Edge computing could be a distributed computing paradigm

that brings computation and data storage closer to where it needed. With the assistance of IoT,
nowadays edge computing is reaching more people. the most aim is to maneuver computer
capabilities and resources closer to the sting of the network.[4]
DRAWBACKS:
 Limited Redundancy
 Longer output time
 Potential loss or corruption of data
This paper aims to implement light associated sign detection exploitation image process techniques
for an autonomous and vehicle. Traffic sign recognition system is utilized to look at traffic signs,
warn a driver and inform sure actions. Quick sturdy and period automatic traffic sign detection and
recognition will assist the driving force and considerably improves driving safety. Automatic
recognition of traffic signs is to boot vital for machine-controlled intelligent driving vehicles. The
experimental result shows the high correct classification of traffic signs pattern with advanced
background pictures additional as a result of the results accomplished in reducing the machine price of
this planned letter.[8]

DRAWBACKS:
 Light intensity faced major problems

3. PROPOSED SYSTEM
Driverless cars have invaded from the automation area with a broad range of capabilities. In this
project, we are using image processing technology. The whole process deals with edge detection and
the Sobel Feldman algorithm. We are using raspberry pi, ultrasonic sensor, servo motor, motor driver,
Logitech camera, power bank in this project.

Fig. 3.1: Block diagram

The block diagram shown below explains the functioning of the project, Raspberry pi plays a key role
in the project which is connected with a web camera, ultrasonic sensor,motor driver and power bank.
This raspberry pi contains a microSD card that has been used to install the programs. Edge detection
algorithm we used in it to detect the road signs is captured by web camera first, based on the
command given by the program installed in the raspberry pi after image processing the prototype
moves according to it. Coming to the motor driver it controls the two set of servomotors which are
used to move the wheels, this motor driver helps to move the servo motors at any direction
independently and the H bridge construction on it helps to provide sufficient power supply to the
servo motors The specifications of our project are given below:
 It detects an obstacle with the help of an ultrasonic sensor which is placed in front of the
prototype. Instead of taking a break suddenly it notifies the distance of the obstacle and
then it will stop when the obstacle gets closer.
 Speed will be raised and decrease based on the speed limit signs, if the road sign indicates

ISSN: 2005-4238 IJAST

Copyright ⓒ 2020 SERSC 6981
 International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp. 6979-6986

40kmph then it will get automatically slow down and if the road sign indicates 100kmph
then it will raise its speed automatically.
 It will move either lift or right turn according to the road sign
 When the speed breakers are detected automatically it slows down the vehicle.
 Path hole recognition is additionally added to this project, it gets stopped when the path
hole detected.

4. SYSTEM ARCHITECTURE
A. RASPBERRY PI
In our project we are using Raspberry PI 3 ModelB. It’s like a small size computer, we can access a
computer monitor without CPU by connecting this, it supports application from word processer,
spreadsheets games and HD videos, we can create a separate media center in it, technical features of
raspberry pi are listed below:

Fig.4.1.1: Raspberry PI
 It consists of 40pin extend excel GPIO to enhance “real world” projects
 The OS that runs the raspberry pi is a raspberry pi OS which is open source. It has a Micro
SD slot for sorting information and loading the operating systems
 The SRAM of this processor is 1GB.
 It contains 2.4GHz of Dual-band and wireless LAN with 5GHz of speed
 Bluetooth 4.2/BLE
 300Mbps of Ethernet speed
 Compare to other processing boards it is cost-effective
 It requires less power, It can operate with 5v power bank

Fig.4.1.2: Pin diagram of raspberry pi

It contains 40 GPIO pins, first 5 pins and 22, 25 pins of raspberry pi are connected with L239D
motor driver. 21, 23, 24 and 39 pins of raspberry pi connected with the ultrasonic sensor. 3 out of 4
USB ports are connected with keyboard, mouse and web camera. HDMI port connects to a computer
monitor with HDMI cable.

ISSN: 2005-4238 IJAST

Copyright ⓒ 2020 SERSC 6982
 International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp. 6979-6986

B. ULTRASONIC SENSOR
An ultrasonic sensor is mainly used for detecting the obstacles. High-frequency sound waves relay on-
premises to provide different echo patterns. It’s a transducer it can transmit and receive the pulse
which reflects when it hits anobstacle around its range. In our project, it helps to detect the obstacle
from a distance and notify the distance of the obstacle when it gets too close it automatically stops the
car and avoid the collisions. It contains four pins Vcc, Trigger, Echo, Ground, these four pins are
connected to the raspberry pi pins and the ground pin of ultrasonic sensor connected to the raspberry
pi ground.

Fig.4.2.1: Ultrasonic sensor

C. LOGITECH CAMERA
Logitech camera is used for capturing images and video footage, it can capture 5mp resolution
images and 1080p HD videos. The fisheye lens present in front of the camera to provide 3D visual.In
our project we are using edge detecting algorithm and Sobel filters for detecting road sign recognition,
this camera monitors the surroundings and detect the road signs, technical specifications of this
camera are listed below:
 Full HD 1080p video calling
 H.264 video compression
 90° view range with 2 omnidirectional mics
 Zooming the image up to 4X and the clarity of the image will be in 1080p
 It consists of Rightlight version 2 technology for the better quality in diverse lighting
environments, even though in a very low light
 Optional camera settings applications for control of pan, tilt, and to zoom
 It can automatically focus on road signs and contains external privacy shutter

Fig.4.3.1: Logitech camera

D. SERVO MOTORS
Servo motor is a simple electric motor. It restrained with the help of servomechanism. These are
mainly used for rotating the actuator. It leads to allow for controlling angle, velocity, and acceleration.
It doesn’t contain any particular class of motor even though servo motor is referred to the motor, this
motor is fitted for closed-loop control system only. It needs a sophisticated controller because a
dedicated module is especially used with servomotors, this can operate both AC and DC servo motors.
It comprises a motor coincident to a sensor with certain positive feedback. Servomotors applications
used in robotic controlling toys etc

ISSN: 2005-4238 IJAST

Copyright ⓒ 2020 SERSC 6983
 International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp. 6979-6986

E. L239D motor driver IC

L239D motor driver is a 16 pin IC, where 3 pins act as Vcc, it contains positive and negative pins,
four positive input and four negative output pins. It can control two sets of servo motors
simultaneously. The gear motor can also be controlled by Arduino, but an Arduino power is
insufficient to drive the motor. It is capable of driving a small and a quiet big motors due to H bridge
construction on it which helps to send the sufficient power supply in any direction, a 5w battery is
connected to it to provide the power supply. It is a typical motor that can rotate the two servo motors
in any direction independently and it can also be used in robotic applications for controlling motors

F. POWER BANK
The power bank is used to provide the power supply, in our project we connected 1000mah power
bank to raspberry pi for power supply.

G. EDGE DETECTION USING SOBEL ALGORITHM

The Sobel operator also called Sobel–Feldman operator or Sobel filter is employed in image
processing and computer vision, mainly within edge detection algorithms where it creates a picture
emphasizing edges. Technically, it's a special distinction operator, computing an approximation of the
gradient of the image intensity function. The Sobel–Feldman operator is predicated on convoluting
the image with alittle, separable, and integer valued filter within the vertical and horizontal directions
and is therefore relatively inexpensive in terms of computations. On the opposite side, the gradient
approximation that it provides is comparatively crude, particularly for the high-frequency variations
within the image

5. FINDINGS AND RESULTS

1. PROJECT SETUP
Below diagram shows the project setup contains raspberry pi, motor driver, servo motors, Logitech
camera, and ultrasonic sensor

Fig.5.1.1: Project setup

2. WORKING RESULTS
The raspberry pi installed with 3 programs using python, one program contains GO, LEFT, RIGHT,
SPEED LIMIT OF 100 & 40, STOP and ULTRASONIC SENSOR commands. PATH HOLE and
SPEED BREAKER contains their separate codes. Fig:5.2.1, Fig:5.2.2 shows the result of speed limit
100, right turning and the readings on the display shows the distance measurement of an object
detected by the ultrasonic sensor. Fig:5.2.3 shows the result of path hole recognition.

ISSN: 2005-4238 IJAST

Copyright ⓒ 2020 SERSC 6984
 International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp. 6979-6986

Fig.5.2.1: Result for a speed limit of 100

Fig.5.2.2: Result for right-turning

Fig.5.2.3: Result for path hole

CONCLUSION
Road sign detection and vehicle controlling without the driver is the main objective of this paper.This
paper explained the system architecture and the process of the project. Ultrasonic sensor fixed in front
of the car, if it gets disturbed with the connections then the program couldn’t run.

ISSN: 2005-4238 IJAST

Copyright ⓒ 2020 SERSC 6985
 International Journal of Advance Science and Technology
Vol. 29, No. 10S, (2020), pp. 6979-6986

REFERENCES
[1] Shun-Yu Chan,” RFID-based autonomous mobile car” Department of Electrical Engineering,
Cheng-Shiu University, Kaohsiung, Taiwan, 978-1-4244-7300-7/10/$26.00©2010 IEEE.
[2] Nivas V M, Gopi Krishnan, Chyrilfredrhic A,” Automated Guided Car(AGC) for Industrial
Automation” UG Scholar, Department of Mechanical Engineering, Panimalar Engineering College,
Chennai, India, 978-1-4673-6725-7/16/$31.00©2016 IEEE.
[3] AmrithanandhaBabu G, Guruvayoorappan K, SajithVariyar V.V, Dr.K.PSoman,” Design and
Fabrication of Robotic Systems: Converting a Conventional Car to a Driverless Car” Centre for
Computational Engineering and Networking(CEN), Amrita School of Engineering, Coimbatore, 978-
1-5090-6367-3/17/$31.00©2017 IEEE
[4] Abdulla FawazAljulayfi, Karim Djemame,” Simulation of an Augmented Reality Application for
Driverless Cars in an Edge Computing Environment”, school of computing, university of Leeds, UK,
978-1-5386-6224-3/18/$31.00©2018 Crown.
[5] Andrew Ydenberg, Navtej Heir, Bob Gill,”Security, SDN, and VANET Technology of Driver-
less cars”, Electrical Engineering, School of Energy, British Colombia Institute of Technology,
Burnaby, Canada, 978-1-5386-4649-6/18/$31.00©2018 Crown.
[6] Xin Zhang, Maolin Chen, Xingqun Zhan,” Behavioral Cloning for Driverless Cars using Transfer
learning”, School of Aeronautics and Astronautics, Shanghai Jiao Tong University, Shanghai, China
978-1-5386-1647-5-/18/$31.00©2018 IEEE.
[7] R. M. Fouad, A. Onsy, Osama A. Omer,” Improvement of Driverless cars passengers on Board
Health and Safety, using Low-Cost Real-Time Heart Rate Monitoring System” School of
Engineering, University of Central Lancashire, Preston, UK, proceedings of the 24 th International
Conference on Automation & Computing, Newcastle University, Newcastle upon Tyne, UK, 6-7
September 2018.
[8] Sandhiyaa S, Vijayan M,” Traffic Light and Detection for Autonomous Land Vehicle Using
Raspberry PI”, Department of ECE, Vellalar College of Engineering and Technology, India,
International Journal of Scientific Research and Review, Volume 07, Issue 03, March 2019.
[9] K.Vinothini, Dr.s.Jayanthy,” Road Sign Recognition System for Autonomous Vehicle using
Raspberry PI” PG Student, Department of ECE, Sri Ramakrishna Engineering College, Coimbatore,
India, 978-1-5386-9533-3/19/$31.00©2019 IEEE
[10] D Jose, PN Kumar, A Hussain, P Shanker,VLSI circuit partitioning using ant colony optimisation
to yield fault tolerant testable systems, Arabian Journal for Science and Engineering 39 (12), 8709-
8729,2014
[11] D Jose, PN Kumar, L Saravakanthan, R, Dheeraj,Computer aided partitioning for design of
parallel testable VLSI systems, IEEE Tnternational Conference on Advances in Computing,
Communications and Informatics (ICACCI), pg. 1363-1366, 2013

ISSN: 2005-4238 IJAST

Copyright ⓒ 2020 SERSC 6986