In summary, the problem of face detection or recognition is alone not sufficient for the security of the cars, and hence eye detection of the same person is required for the system to be highly secured to give access for the car for owner or driver and deny for the intruder. A Review of journal and conference papers on face detection, recognition, eye detection, driver fatigue detection are done to summarise the techniques or methods employed, along with the outcomes and drawbacks or disadvantages. As an intruder may not manifest all symptoms of fatigue using a single detection system, a hybrid system is required to integrate the result of different systems and detect the intruder or to recognise the owner or driver with higher accuracy. When the result from both systems are positive, the hybrid system will determine that the car owner or driver is true with a high accuracy.
Table of Contents
1. INTRODUCTION
1.1 Introduction
1.2 Research problem
1.3 Aim and objectives
1.4 Justification for this research
1.5 Organization of the rest of the chapters
1.6 Summary
2. LITERATURE REVIEW
2.1 Introduction
2.2 Literature review
2.2.1 Facial Detection Methods
2.2.2 False Statistical Methods
2.2.3 Image Processing Methods
2.2.4 Drawbacks and disadvantages of the existing methods
2.3 Summary
3. CONCEPT DESIGN AND RESEARCH METHODOLOGY
3.1 Introduction
3.2 Investigation on material and component selection
3.2.1 Pre-processing
3.2.2 Software selection
3.3 Proposed methodology
3.4 Concept design based on fundamental engineering principles
3.4.1 Facial detection system
3.4.2 Perclos
3.4.3 Thresholding
3.4.4 Iris Recognition
3.4.5 GUI
3.5 Professional engineering practices
3.6 Summary
4. FINAL DESIGN AND SYSTEM IMPLEMENTATION
4.1 Introduction
4.2 System Implementation
4.2.1 Facial Detection
4.2.2 Image Processing
4.2.3 Feature Isolation
4.2.4 Eye State Evaluation
4.2.5 Mouth State Evaluation
4.2.6 Security Detection and Classification
4.2.7 Alarm System
4.3 Working Principle
4.4 Results
5. PROJECT FINDINGS AND TESTING
5.1 Testing of the Proposed Design
5.1.1 Classifier Detection test
5.1.2 Face Detection Test 2
5.1.3 CLAHE Test
5.1.4 System Speed Test
5.1.5 State Evaluation Test
5.2 Discrepancy between theoretical and experimental
5.3 Error and Troubleshooting
5.4 Comparison with Prior Researches
5.5 Sustainable Developemnt and Environmental
5.6 Project Management, Finance and Entrepreneurship
5.6.1 Project Management
5.6.2 Finance
5.6.3 Entrepreneurship
5.7 Moral Professionalism and Ethical consideration
5.8 Contribution of this project
5.8.1 Hybrid Cascade Classifier
5.8.2 Smaller Face Region Detection
5.8.3 Alarm System
5.8.4 Eye State Detection
6. CONCLUSION AND RECOMMENDATION
6.1 Conclusion
6.2 Limitations
6.2.1 Facial Detection Limitations
6.2.2 Security Detection
6.3 Recommendations
6.3.1 Facial Detection
6.3.2 Security Detection Method
6.3.3 Further Implementations
6.4 Summary
Project Goals and Key Topics
The primary aim of this project is to enhance vehicle security by developing a hybrid detection system that integrates advanced facial recognition with eye state detection. By moving beyond single-mode detection, the project seeks to provide a highly accurate authentication and fatigue-monitoring solution that discriminates reliably between the car owner/driver and potential intruders.
- Development of a combined facial recognition and eye detection system utilizing image processing and MATLAB.
- Implementation of a high-security user authentication mechanism to prevent unauthorized vehicle access.
- Design of a robust Graphical User Interface (GUI) for system interaction and verification.
- Evaluation of system accuracy against diverse environmental conditions and subject demographics.
- Analysis of fatigue detection parameters, including PERCLOS and yawning-based indicators.
Excerpt from the Book
1.1 Introduction
Vehicle security system is gaining its popularity in the recent years and and vehicles can be as much as intelligent using the internet of things. However, these vehicles suffer from various lot of crimes. Hence, it is a very gig challenge to overcome them. Most of the vehicles are so controlled with security cars, password or mechanical based keys. Any of these improvements just depend on the owner and responsibility is on them (Elangar & Kayed, 2020)
Generally, the intruder behaviour and alertness of drivers or car owners could be determined in several ways, i.e. physiological, psychological, physical. Physiological can be defined as the inner bodily reactions related to the transportation system and nervous system such as heart rate (HR), respiration rate (RR), and heart rate variance (HRV), other methods include brainwaves or eye movement detection via various electrography.
Psychological detection method is done via monitoring natural bodily reactions such as jerking, yawning, blinking, Percentage of Eyelid Closure over the Pupil over Time (PERCLOS), and other involuntary movements. Physical can also be classified as motor detection, related to the voluntary bodily movements such as steering input, braking, accelerating, and gear shifting, vehicle yaw angler variation among others. Some systems may incorporate several key detection points at once to increase accuracy, such as yawning and blinking for vision detection systems, or braking and steering for motion detection systems.
The ideal system should be non-contact, low in latency, highly accurate, and convenient. This means that the system should have a steady response rate so as to maintain continuous operation along the driving session, ensuring detection to be effective throughout. It should be accurate, with little to none false positives or ignored positives, so that the detected data has high fidelity compared to the reality of the driver’s condition, and the alarm timings would be appropriate. Finally, no setup should be required on the driver’s part, the system should be fully automated and operational once the vehicle is intended to be started to ensure no errors due to setup can occur.
Summary of Chapters
1. INTRODUCTION: This chapter introduces the growing necessity for intelligent vehicle security systems and outlines the project's specific objectives regarding facial and eye detection accuracy.
2. LITERATURE REVIEW: This chapter reviews contemporary research from 2015 onwards, analyzing various facial and eye detection techniques, their methodologies, and identified shortcomings in existing systems.
3. CONCEPT DESIGN AND RESEARCH METHODOLOGY: This chapter details the core architecture of the proposed hybrid system, including pre-processing techniques, software selection, and fundamental engineering principles used to build the detection pipeline.
4. FINAL DESIGN AND SYSTEM IMPLEMENTATION: This chapter describes the practical realization of the security system, explaining the implementation of Haar Cascade and LBP classifiers, image processing steps like CLAHE, and the integration of an alarm system.
5. PROJECT FINDINGS AND TESTING: This chapter presents the comprehensive evaluation of the developed system across multiple test scenarios, focusing on detection rates, robustness to varying lighting, and system speed.
6. CONCLUSION AND RECOMMENDATIONS: This chapter summarizes the project's achievements, acknowledges specific technical limitations regarding lighting and angle variations, and suggests future improvements like angle-invariant detection.
Keywords
vehicle security, facial recognition, eye detection, PERCLOS, image processing, MATLAB, Haar Cascade, LBP classifier, hybrid system, driver fatigue, authentication, CLAHE, security detection, algorithmic robustness, system implementation
Frequently Asked Questions
What is the core purpose of this research project?
The project focuses on developing a high-security vehicle access and driver monitoring system that uses both facial and eye detection to reliably differentiate between authorized owners and intruders.
Which central thematic areas are explored?
Key areas include the integration of hybrid classification methods, image processing for feature extraction under varying illumination, and the implementation of real-time fatigue monitoring.
What is the primary research objective?
The primary aim is to design a high-accuracy, non-intrusive security system capable of recognizing a driver's face and monitoring for symptoms of fatigue or unauthorized access.
Which scientific methodology is employed?
The research uses a hybrid approach combining Haar Cascade and Local Binary Pattern (LBP) classifiers, supplemented by image processing techniques like Contrast Limited Adaptive Histogram Equalization (CLAHE) for feature enhancement.
What is the focus of the main body of the work?
The main body centers on the system architecture, mathematical formulations for eye and mouth state evaluation, and experimental validation of the software implementation in Python/MATLAB.
Which keywords best describe this study?
Significant keywords cover vehicle security, facial recognition, eye state analysis, algorithmic efficiency, and hybrid classifier integration.
How does this system handle low-light environments?
The system utilizes CLAHE to equalize images, though the author notes that lighting remains a challenge and suggests future integration of IR cameras for night vision.
Why are both facial and eye detection necessary?
The research concludes that facial detection alone is insufficient, and integrating eye detection provides the redundancy and accuracy required to prevent intruder access and identify fatigue effectively.
What role does the Graphical User Interface (GUI) play?
The GUI provides an interface for image acquisition and real-time authentication feedback, enabling the user to monitor system status and verification results.
- Arbeit zitieren
- Bandar Hezam (Autor:in), 2021, Development of High Security System for Cars, München, GRIN Verlag, https://www.grin.com/document/1357953
