Scribd
Upload
2 views

Quantum_Machine_Learning_for_EEG_Based_Seizure_Detection__Quantum_GANs_and_Quantum_Boltzmann_Machines

This paper presents an intelligent traffic signal control system that utilizes YOLO for real-time vehicle detection and an OV7670 camera integrated with an Arduino microcontroller to optimize traffic flow. The system dynamically adjusts signal timings based on real-time traffic density, achieving a vehicle detection accuracy of 94.5% and reducing average waiting times by 38% compared to traditional fixed-timer signals. Future enhancements include the deployment of edge AI for processing efficiency and IoT-based traffic monitoring for scalability.

Uploaded by

Palnati Datta
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
2 views

Quantum_Machine_Learning_for_EEG_Based_Seizure_Detection__Quantum_GANs_and_Quantum_Boltzmann_Machines

This paper presents an intelligent traffic signal control system that utilizes YOLO for real-time vehicle detection and an OV7670 camera integrated with an Arduino microcontroller to optimize traffic flow. The system dynamically adjusts signal timings based on real-time traffic density, achieving a vehicle detection accuracy of 94.5% and reducing average waiting times by 38% compared to traditional fixed-timer signals. Future enhancements include the deployment of edge AI for processing efficiency and IoT-based traffic monitoring for scalability.

Uploaded by

Palnati Datta
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.

Digital Object Identifier 10.1109/ACCESS.2024.0429000

Real-Time Intelligent Traffic Signal Control Using


YOLO and OV7670 Camera with Arduino
Integration
M.SAMBATH, REPAKA SAI AKSHITH, P.V.S. UDAY KIRAN, I.SANTHOSH REDDY
Vellore Institute of Technology (VIT), Chennai, Tamil Nadu, India (e-mail: [email protected])
Corresponding author:m.sambath (e-mail:[email protected] ).

ABSTRACT
bullet.png
Traffic congestion is a critical problem in urban settings, leading to delays, increased fuel consumption,
and elevated emissions. Traditional traffic signal control relies on fixed timers, which are inefficient in
adapting to changing traffic conditions. This paper proposes an intelligent traffic signal control system
that utilizes YOLO for real-time vehicle detection and an OV7670 camera integrated with an Arduino
microcontroller. The system captures real-time traffic images, evaluates vehicle density, and dynamically
adjusts the green light duration to optimize traffic flow. Experimental results demonstrate significant
improvements in reducing wait times and enhancing traffic efficiency compared to conventional traffic
control systems.

INDEX TERMS Traffic management, YOLO, object detection, real-time systems, Arduino, OV7670, deep
bullet.png
learning.

I. INTRODUCTION signal timings based on real-time traffic density.


Traffic congestion is a major challenge in urban areas, causing
economic losses and environmental degradation. The increase II. SYSTEM ARCHITECTURE
in vehicle numbers and inefficient traffic signal systems ex- The proposed system comprises three key components:
acerbate this issue. Traditional traffic control systems rely on • OV7670 Camera Module: Captures real-time traffic
preset signal timers, which do not adapt to varying traffic images and transmits them for processing.
conditions, leading to unnecessary delays and increased fuel • YOLO-Based Object Detection: Detects and classifies
consumption. vehicles from captured images.
Modern solutions incorporate sensors and computer vision • Arduino-Based Traffic Light Controller: Adjusts sig-
techniques for adaptive traffic management. Sensor-based nal timing dynamically based on traffic density.
systems, such as infrared and inductive loop sensors, are The system captures images at intersections, processes
costly and affected by environmental factors. Conversely, them using the YOLO model to detect and count vehicles, and
image-processing-based systems leverage computer vision adjusts traffic signals accordingly. A serial communication
techniques to analyze traffic conditions in real time. However, link between the Arduino and the processing unit enables
traditional methods such as edge detection and contour analy- real-time control of signal durations.
sis are susceptible to inaccuracies due to lighting changes and
occlusions. III. EXPERIMENTAL RESULTS
Deep learning approaches have proven highly effective for The system was tested under various traffic conditions to
real-time traffic analysis. YOLO (You Only Look Once) is an evaluate performance metrics such as detection accuracy,
advanced object detection framework capable of identifying processing time, and traffic flow improvement.
multiple objects in a single forward pass, making it ideal • Detection Accuracy: YOLO achieved a vehicle detec-
for real-time applications. This research integrates YOLO tion accuracy of 94.5%.
with an Arduino-controlled OV7670 camera to develop an • Processing Speed: The system processed each frame in
adaptive traffic signal control system that dynamically adjusts 15ms, ensuring real-time operation.

VOLUME 11, 2023 1


Joseph Redmon et al.: Real-Time Intelligent Traffic Signal Control Using YOLO and OV7670 Camera

• Traffic Efficiency: The adaptive system reduced av-


erage waiting times by 38% compared to fixed-timer
signals.

IV. CONCLUSION AND FUTURE WORK


This paper presents an intelligent traffic control system that
integrates YOLO and an Arduino-controlled OV7670 camera
to dynamically optimize signal timings. The experimental
results demonstrate significant improvements in traffic ef-
ficiency. Future enhancements include deploying edge AI
for processing efficiency and incorporating IoT-based traffic
monitoring for scalability.

REFERENCES
[1] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You Only Look Once:
Unified, Real-Time Object Detection," in Proc. IEEE CVPR, 2016.
[2] M. Everingham et al., "The PASCAL Visual Object Classes Challenge,"
Int. J. Comput. Vis., vol. 111, no. 1, pp. 98–136, 2015.
[3] S. Ren, K. He, R. Girshick, and J. Sun, "Faster R-CNN: Towards Real-Time
Object Detection with Region Proposal Networks," Neural Inf. Process.
Syst., 2015.
[4] A. Bochkovskiy, C.-Y. Wang, and H.-Y. M. Liao, "YOLOv4: Op-
timal Speed and Accuracy of Object Detection," arXiv preprint
arXiv:2004.10934, 2020.
[5] T.-Y. Lin et al., "Feature Pyramid Networks for Object Detection," Proc.
IEEE CVPR, 2017.

2 VOLUME 11, 2023

You might also like