Blind spot accidents are a common cause of vehicle collisions, particularly during lane changes or merging into traffic. The Vision-Based Speed Detection in a Blind Spot Monitoring System addresses this issue by providing a real-time solution for monitoring blind spots, detecting objects, and estimating their speed to prevent potential accidents. This system utilizes an IP camera mounted strategically on the vehicle, such as on the side mirrors or rear bumper, to capture video footage from areas that are typically outside the driver’s field of view. The captured video feed is processed using advanced computer vision techniques powered by the YOLOv8 object detection model. The system identifies objects such as vehicles, motorcycles, or pedestrians in the blind spot and tracks their movements. Through frame-by-frame analysis, it calculates the speed of the detected objects and determines whether they pose a collision risk based on their proximity and velocity. If a hazard is detected, the system immediately triggers an audible alarm, alerting the driver to take appropriate action. The system is designed for real-time implementation in a vehicle environment. Unlike traditional systems that require external devices like laptops for processing, this paper is intended to be embedded within the vehicle using compact and cost-effective hardware, such as a Raspberry Pi or equivalent processing units. The alarm system integrates with the vehicle’s sound system or operates as an independent module to provide clear and immediate warnings. This paper represents an innovative approach to vehicle safety by combining object detection, speed estimation, and hazard assessment into a single cohesive system. Its real-time functionality ensures drivers are equipped with critical information when they need it most, enhancing road safety and reducing the likelihood of collisions. Moreover, the modular nature of the system allows for scalability, such as integration with existing Advanced Driver Assistance Systems (ADAS), enabling further enhancements like lane departure alerts or adaptive cruise control. The Vision-Based Speed Detection in a Blind Spot Monitoring System is a step toward smarter and safer transportation. By leveraging cutting-edge computer vision technology, the paper not only addresses a critical road safety issue but also paves the way for future innovations in autonomous and semi-autonomous vehicles.
Blind Spot Monitoring and Detection System using Computer Vision
28.04.2025
547983 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Blind-spot monitoring using machine vision and precise FOV information
Europäisches Patentamt | 2019
|Blind-spot monitoring using machine vision and precise FOV information
Europäisches Patentamt | 2018
|Blind-spot monitoring using machine vision and precise FOV information
Europäisches Patentamt | 2020
|Vehicle blind spot monitoring system and vehicle blind spot monitoring method
Europäisches Patentamt | 2024
|