Joginpally B.R Engineering College: Avoid Road Accidents Using Macine Learning Algorithm

You might also like

Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 32

JOGINPALLY B.

R ENGINEERING COLLEGE
A MICRO PROJECT REPORT ON

AVOID ROAD ACCIDENTS USING MACINE LEARNING ALGORITHM

UNDER THE GUIDENCE OF PRESENTED BY :

MR.M.L.M PRASAD 1. THELLA VEERAPRATHAP


2. VANGA SOUMYA
3. VASAM VAMSHI
4. MOGILI SAI RAM
TABLE OF CONTENTS
 Abstract
 Introduction
 Existing system
 Proposed system
 Modules
 Software requirements specification
 System design
 System implementation
 Testing
 Testing strategies
 Test cases
 Screenshots
 Conclusion
 References
ABSTRACT
 Fatigue among drivers is a major cause of road accidents every year in India. Lack
of sound sleep for six to eight hours is one of the primary reasons behind this
fatigue.
 Drivers with sleep deprivation can imbalance the reaction time and decision making
when behind the wheels and this can increase the cause of accidents. It may cause
death or severe injuries.
 Therefore, it is highly essential to create a smart system that can spot and alert the
driver of his/her condition. Although there are few solutions proposed in this
direction, most of them have not been implemented successfully and many of them
only remain in theory.
 In this research paper, we propose an efficient driver fatigue detection and alert
system using mainly open-source technologies. We implement and test this system
in real-time and the results are highly encouraging compared to many existing
systems.
INTRODUCTION

 Every day we see there are many road accidents occurring on the highways causing many
deaths. The drivers falling asleep while driving is a silent killer.
 Many may not notice, but the accidents caused by drivers falling asleep contribute nearly
40% of road accidents in India.
 The reason drivers fall asleep is mainly due to fatigue and long working hours.
 The main motive is to build a system which detects sleep and raises the alarm .
 To detect the human body are tracked with various parameters of facial features.
 Factors causing the accidents speeding, drunk and drive, reckless driving, night driving .
Here we are concerned about the factors which make drivers fall asleep.
EXISTING SYSTEM
 Earlier solutions are all based on 3D face-mapping technology and tracking and analyzing the
face based on the image.
 Few systems which are based on fig-net database were able to achieve good detection rates,
but they have high false alarm rates.
 There are also systems which use the techniques of the supervised and unsupervised linear
transformation and the cascade detection methods systems which managed to achieve low
false alarm rate are not so efficient in detecting.

Disadvantages of Existing System


 Low accuracy in detection
 Very high false alarm rate
PROPOSED SYSTEM
 The proposed application is based on the technology of computer vision and machine
learning.
 A small camera embedded in the device feeds into a processor, which identifies the
driver’s face and eyes and measures his facial patterns at the time he is drowsy or is about
to fall asleep.
 Thus, this new technology is able to send out timely alerts or warn the driver to prevent
the occurrence of any accidents.
ADVANTAGES OF PROPOSED SYSTEM
 User-friendly and cost-effective
 Helps in gathering information on driving habits of individuals
 Better results comparing to the existing systems in both detection and false alarm rate
 Enhancing connectivity on highways due to a reduced number of accidents causing
traffic jams.
MODULES

 Camera module
 Eye and mouth detection
 Scoring
 Measure tracking
Camera module
 The job of the camera module is to activate the camera and use the classifiers to evaluate your
features.
 The system will keep taking a picture until a person is detected. Once a face is detected, we will
calibrate the features. This will produce the following initial ratio calculations.
Eye and mouth detection
 Given the image taken by the camera, the images are sent for the eye and mouth detection module.
 Here in this module, the classification module is run. Here the eyes are detected by the ratios, while
mouth detection is done by detecting the lips position
Scoring
 Here in the scoring module, the score is given for the images after detecting their face.
 Scoring will be done by classifying the images based on eye and face detection. The naive Bayes
method is used for scoring.
Measure tracking
 This module's purpose is to evaluate the images from both eye and mouth detection, after classifying
the sleepiness is detected. Once it detects the drowsiness, it raises an alarm to wake up the driver.
SOFTWARE REQUIREMENT SPECIFICATION

HARDWARE REQUIREMENTS
 Processor : i5 processor
 Hard disk: 258MB
 RAM : 4GB

SOFTWARE REQUIREMENTS
 Operating system : Mac OS
 Framework : Open cv, Tensor flow, Scikit –learn
 Technologies : Python
 IDE : VS code, Googlecollab
SYSTEM DESIGN
System design is the process of designing the elements of a system such as the architecture, modules, and
components, the different interfaces of those components, and the data that goes through that system.

ELEMENTS OF A SYSTEM
▪ Architecture - this is the conceptual model that defines the structure, behavior, and more views of a system.
We can use flowcharts to represent and illustrate the architecture.
▪ Modules - these is components that handle one specific task in a system. A combination of the modules
makes up the system.
▪ Components - this provides a particular function or group of related functions. They are made up of
modules.
▪ Interfaces - this is the shared boundary across which the components of a system exchange information and
relate.
▪ Data - this the management of the information and data flow. Mini tasks performed during the system design
process.
UML DIAGRAMS
 UML is a way of visualizing a software program using a collection of diagrams.
 Today, UML is accepted by the object management group (OMG) as the standard for modeling
software development. UML stands for unified modeling language.
 Improved integration between structural models like class diagrams and behavior models like
activity diagrams.
 Added the ability to define a hierarchy and decompose a software system into components and sub-
components.
Class Diagram
 A class diagram is a static diagram.
 It represents the static view of an application.
Use Case Diagram
 A use case diagram is a dynamic or behavior diagram in UML.
 Use case diagrams model the functionality of a system using actors and use cases
SEQUENCE DIAGRAM
Collaboration Diagram
ACTIVITY DIAGRAM
 Activity diagram is another important behavioral diagram in UML diagram to describe dynamic aspects
of the system.
 Activity diagram is essentially an advanced version of flow chart that modeling the flow from one
activity to another activity.
State chart Diagram
 State chart diagram is one of the five UML diagrams used to model the dynamic nature of a system.
 They define different states of an object during its lifetime and these states are changed by events.
 State chart diagrams are useful to model the reactive systems.
DEPLOYMENT DIAGRAM
 Deployment diagrams are important for visualizing, specifying, and documenting embedded,
client/server, and distributed systems and also for managing executable systems through forward and
reverse engineering.
 A deployment diagram is just a special kind of class diagram, which focuses on a 20 system's nodes.
 Graphically, a deployment diagram is a collection of vertices and arcs. Deployment diagrams commonly
contain:
COMPONENT DIAGRAM
 A component diagram breaks down the actual system under development into various high levels of
functionality.
 Each component is responsible for one clear aim within the entire system and only interacts with other
essential elements on a need-to-know basis.
SYSTEM IMPLEMENTATION
System Architecture
PYTHON
Python is a general-purpose, versatile, and powerful programming language.
The syntax of the language is clean and length of the code is relatively short.

 It’s a great first language because it’s concise and easy to read.
 Whatever you want to do, python can do it. From web development to machine learning to data science,
python is the language for you.
 Great first language
 Large programming community
 Excellent online documentation
 Endless libraries and packages
 World-wide popularity
 Powerful and flexible
MACHINE LEARNING

 Machine learning is a data analytics technique which gives


computers the capability to perform specific tasks without
any explicit instructions .
 The process of learning is done through observations and
data.
 The primary aim of this technique is to allow computers to
learn automatically without human intervention or
assistance .
SYSTEM TESTING

System testing is a level of software testing where a complete and integrated software is tested. The purpose of
this test is to evaluate the system's compliance with the specified requirements.

 The process of executing a system with the intent of finding an error.


 Quality is defined as justification of the requirements.
 Testing can demonstrate the presence of bugs, but not their absence.
 Debugging and testing is not the same thing.
 Testing is a systematic attempt to break a program or the AUT.
TESTING STRATEGIES
Black box testing: it is the testing process in which tester can perform testing on an application without
having any internal structural knowledge of application. Usually test engineers are involved in the black box
testing.
White box testing: it is the testing process in which tester can perform testing on an application with having
internal structural knowledge. Usually, the developers are involved in white box testing.
Levels of testing :levels of testing includes system testing, unit testing, integration testing and acceptance
testing.
Unit testing :in unit testing, we designed the whole system in a modularized pattern and each module was
tested. Till we get the accurate output from the individual module we worked on the same module.
Integration testing : after constructing individual modules all the modules were merged and a complete
system was made. Then the system was tested whether the system was producing results correctly and the
produced results were accurate or not.
TEST CASES
FINAL TABLE
Test case No.of Observations No.of hits Accuracy
Yawn Detection(mouth) 15 12 80%
Eye Detection 15 15 100%
Assent Head Right 15 14 93%
Assent Head Left 15 14 93%
Nodding 15 14 93%

Overall Accuracy
92%
CONCLUSION
 The system with high accuracy and good alarm rate which are not being able
to exist simountensly are now implemented within one system, also with the
help of statistical inference and advanced image detection algorithms an
accuracy of 88% is achieved.
 The system is capable of accurately detecting sleep, using the parameters
like eye and mouth states and their position we are able to effectively detect
the driver's fatigue status.

Future Enhancement
 Despite achieving great accuracy there are still few flaws left in our system. One flaw of this
application is the bias. Unfortunately, for some different facial features, the system will go crazy
in detecting the lips and the eyes.
REFERENCE

 Https://www.Researchgate.Net/publication/271376030_real_time_drowsy_driver_
detection_using_haarcascade_samples
 https://ieeexplore.Ieee.Org/abstract/document/8748448
 J. L. Crowley, T. Cootes, “FGNET, face and gesture recognition working
group”,http://wwwprimainrialpes.Fr/fgnet/html/home.Html, (2009)
 Arulkumar, C. V., And P. Vivekananda. "Multi-feature based automatic face identification on kernel Eigen
spaces (KES) under unstable lighting conditions." Advanced computing and communication systems, 2015
international conference on. IEEE, 2015
 Vandna Saini, Rekha Saini “driver drowsiness detection system and techniques”, IJCSIT, vol. 5 (3)

You might also like