1. 07/28/2025 1
KOFORIDUA TECHNICAL UNIVERSITY
FACULTY OF ENGINEERING
DEPARTMENT OF AUTOMOTIVE ENGINEERING
COURSE: RESEARCH METHODOLOGY II
TOPIC:
MAKING ROAD SIGNS RECOGNIZABLE BY VEHICLES
2. 07/28/2025 2
GROUP MEMBERS
Ahortor Asraku Sampson
Antwi Isaac
Yeboah Clement
Obeng Oscar
- B301220029
- B301220008
- B301220018
- B301220010
3. 07/28/2025 3
ARTICLE 1
An automated system for traffic sign Recognition using Convolutional
Neural Network
BACKGROUND
PROBLEM STATEMENT
• Manual inventory is laborious; prior methods lack
real-time robustness.
• Drivers may miss signs at high speed
SOLUTION
• End-to-end CNN on 43 classes; 95% accuracy via
TensorFlow implementation.
4. 07/28/2025 4
ARTICLE 2
Traffic Sign Detection for Intelligent Transportation Systems.
BACKGROUND
PROBLEM STATEMENT
• Variability in sign appearance and environment
impedes detection.
SOLUTION
• Color-based, shape-based, ML/DL methods; call
for unified benchmarks.
5. 07/28/2025 5
ARTICLE 3
Analysis of Market-Ready Traffic Sign Recognition Systems in
Cars: A Test Field Study
BACKGROUND
PROBLEM STATEMENT
• Cross-brand variability in hardware/software
causes inconsistent accuracy.
SOLUTION
• Standardize TSRS benchmarks and improve sign-
design tolerances.
6. 07/28/2025 6
INSTITUTION Koforidua Technical University
PROGRAMME B’Tech Automotive Engineering
FULL NAME Clement Kwadwo Yeboah
INDEX NUMBER B301220018
TOPIC Traffic Sign Recognition By Vehicles
7. 07/28/2025 7
What is Traffic Sign Recognition?
A technology that enables vehicles to detect, classify, and respond to road signs using visual sensors (cameras) and
artificial intelligence (AI).
How Traffic Sign Recognition Works
1. Eyes: Cameras that detect and capture sign images.
2. Brain: AI tools like CNN or YOLO identify the sign type.
3. Action: Vehicle adjusts speed or direction accordingly.
Why Do We Need It?
4. Safety: Prevents accidents.
5. Compliance: Adheres to traffic laws.
6. Autonomy: Essential for self-driving cars
These are true because the ASTJ-2025 research shows that during clear weather conditions, the YOLOv8 Nano shows
99.5 accuracy in labs while for humans, 98%.
8. 07/28/2025 8
The Core Challenge
Autonomous vehicles struggle with real-time accurate traffic sign recognition due
to a few noted reasons. Some of them are:
• Environmental factors: Fog or mist, lighting and snow.
• Occlusions: Obstructed signs.
• Cost versus accuracy: High performance systems are expensive.
• Real-time demands: Must process signs in less than 100ms to be useful.
Some of these factors cause a drop in the accuracy of visual sensors and
thereby cause 23.7% of sign related accidents.
9. 07/28/2025 9
Conclusion
Traffic sign recognition isn’t just about teaching cars to see—it’s
about building systems that adapt when vision fails. As these papers
show, the future lies not in choosing one approach, but in merging
their strengths to create roads that are safer for everyone.
