ISSN:2582-5208

www.irjmets.com

Paper Key : IRJ************053
Author: Bhavesh Anil Pawar
Date Published: 21 Oct 2023
Abstract
In recent years, there has been a growing interest in sports and the importance of video recording and analysis. This is particularly relevant in sports like soccer, where complex and fast events occur. Ball detection and tracking, as well as player analysis, have become crucial tools for coaches in assessing team performance and making optimal decisions. Video analysis is also used by recruiters to identify talented players based on their previous games. Ball detection is valuable in helping referees make important decisions during crucial moments of the game.To address the challenges of ball detection and player tracking in soccer videos, researchers have proposed a deep learning-based model called YOLOv3. This model utilizes advanced techniques to detect and track the ball and players in broadcast soccer videos. Initially, the videos are processed to remove unnecessary parts like zoom-ins and replays, ensuring that only relevant frames are used for analysis. The tracking is achieved using the SORT algorithm, which utilizes Kalman filtering and bounding box overlap to accurately track the ball and players throughout the game.Overall, this approach enables coaches, recruiters, and referees to have a better understanding of the game by analyzing the movements of the ball and players in soccer videos.
Paper File to download :