Collision avoidance systems are vital for improving the safety and efficiency of autonomous vehicles, particularly in preventing accidents caused by human errors such as delayed reactions, incorrect time-to-collision (TTC) estimation, and poor judgment during overtaking maneuvers. As the field of autonomous driving continues to evolve, there is a growing demand for fully automated vehicles capable of intelligently interacting with their environment to minimize human intervention while ensuring comfort and safety. In this study, we present a comprehensive collision avoidance system that makes use of cutting-edge computer vision techniques, such as Canny edge detection for lane detection and the YOLOv4 model for real-time object detection and distance calculation. Canny edge detection guarantees accurate lane recognition, which is essential for safe navigation in a variety of driving settings, while YOLOv4 improves the system’s capacity to precisely detect, classify, and track objects. The technology also uses sensor fusion and deep learning to forecast how nearby objects would behave. Distance estimation is used to generate commands, and warnings like “danger, you are going to collide” are sent out to cause quick remedial action. The system’s performance is assessed both during the day and at night in a variety of driving situations, including highway, rural, and urban settings. Accuracy measures for lane detection, distance estimates, and object identification are used to gauge effectiveness. High accuracy in object identification, distance estimation, and lane recognition is demonstrated by the results, allowing for prompt danger detection and command creation to prevent any collisions. In order to create a dependable and expandable collision avoidance system for autonomous cars, this study emphasizes the need of combining YOLOv4, Canny edge detection, and command generation.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Collision Avoidance System in Autonomous Vehicles Using YOLOv4


    Contributors:


    Publication date :

    2024-12-05


    Size :

    621022 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Collision-avoidance system for autonomous-capable vehicles

    GRAY ANDREW | European Patent Office | 2020

    Free access

    Pedestrian Collision Avoidance System for Autonomous Vehicles

    Watzenig, Daniel / Schratter, Markus / Hartmann, Michael | SAE Technical Papers | 2019


    - COLLISION-AVOIDANCE SYSTEM FOR AUTONOMOUS-CAPABLE VEHICLES

    European Patent Office | 2019

    Free access

    COLLISION-AVOIDANCE SYSTEM FOR AUTONOMOUS-CAPABLE VEHICLES

    GRAY ANDREW | European Patent Office | 2024

    Free access

    COLLISION-AVOIDANCE SYSTEM FOR AUTONOMOUS-CAPABLE VEHICLES

    GRAY ANDREW | European Patent Office | 2021

    Free access