Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{182670,
        author = {Sushant R. Upadhyay and Shubhangi Tidake and Prashant Kulkarni},
        title = {Real-Time Detection of Fast-Moving Objects in Dynamic Environments Using YOLO-Based Architectures},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {2},
        pages = {3171-3175},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=182670},
        abstract = {This project addresses the challenge of detecting fast-moving objects—specifically tennis balls in match footage—using deep learning techniques. Traditional object detection models often struggle with motion blur, low latency, and dynamic backgrounds, resulting in decreased accuracy. To overcome these limitations, this work employs YOLO-based architectures, particularly YOLOv8, with focused modifications for high-speed object detection.
A diverse dataset was prepared and augmented with transformations such as hue shifts, brightness changes, mosaic and flip techniques, and simulated motion blur. Extensive hyperparameter tuning and model refinement were conducted to increase robustness and frame-to-frame consistency. The model was then evaluated across videos at various playback speeds to test its real-time performance.
Results showed a 20–30% improvement in detection accuracy compared to baseline, with enhanced performance in low and moderate-speed videos. Novel techniques such as bounding box interpolation and hit detection logic were also developed to track shots and estimate ball speed. These innovations provide practical insights for real-time sports analytics, setting the foundation for future enhancements in ball tracking and trajectory prediction.},
        keywords = {},
        month = {July},
        }