Wrong-Rote-VehicleDetectionSecurity-Alert
This project aims to enhance road safety by detecting vehicles taking wrong routes. When a violation is detected, alerts are sent to the nearby traffic control room and the vehicle owner.
Created on 15th August 2024
•
Wrong-Rote-VehicleDetectionSecurity-Alert
This project aims to enhance road safety by detecting vehicles taking wrong routes. When a violation is detected, alerts are sent to the nearby traffic control room and the vehicle owner.
The problem Wrong-Rote-VehicleDetectionSecurity-Alert solves
Wrong-way driving is one of the primary causes of traffic jams and road accidents. The accurate detection of wrong routes in real-time, combined with prompt security alerts, can significantly enhance safety measures. Our research introduces a solution leveraging You Only Look Once version 8 (YOLOv8) and centroid tracking algorithm, along with Optical Character Recognition (OCR) techniques, for real-time identification of wrong-way vehicles and generation of security alerts.It utilizes YOLO v8 for real-time vehicle detection and Optical Character Recognition (OCR) to read number plates.
Features
Real-Time Vehicle Detection: Utilizes YOLO v8 for accurate and efficient vehicle detection.
Number Plate Recognition: Uses OCR to extract the number plate information from detected vehicles.
Alert System: Sends alerts to nearby traffic control rooms and the vehicle owner in case of a violation.
Project Structure
1.)Video files/ - All datasets for testing
2.)yolo/ - Pre-trained YOLO v8 models
3.)ocr/ - OCR-related scripts and models
4.)requirements.txt - List of dependencies
5.)README.md - Project documentation
Challenges I ran into
1.)Data Quality and Annotation Issues:
Bug/Hurdle: The dataset used for training might have incorrect or inconsistent annotations, affecting the model's accuracy.
2.)Model Performance:
Bug/Hurdle: The vehicle detection model might not perform well in diverse or challenging environments (e.g., varying lighting conditions, occlusions).
3.)Real-Time Processing Constraints:
Bug/Hurdle: Achieving real-time vehicle detection and security alerts can be computationally intensive and may lead to performance bottlenecks.
4.)Integration with Security Systems:
Bug/Hurdle: Integrating the detection system with existing security infrastructure (e.g., alarms, cameras) might present compatibility issues.
5.)False Positives/Negatives:
Bug/Hurdle: The system may generate false alerts or fail to detect vehicles accurately, leading to reliability issues.
6.)Scalability and Maintenance:
Bug/Hurdle: Scaling the system for broader deployment or maintaining it over time could be challenging.
Future Scope
1.Application of system in rural areas
2.Application of system in too busy area with too much traffic
3.Application of gaze estimation and emotion analysis to predict accidents and cause of wrong route movements
Tracks Applied (1)
DevOps
Cheer Project
Cheering for a project means supporting a project you like with as little as 0.0025 ETH. Right now, you can Cheer using ETH on Arbitrum, Optimism and Base.