Virtual Umpiring and Ball Tracking System by Image
Processing using Resnet-18
Journal:
GRENZE International Journal of Engineering and Technology
Authors:
D.Hemanathan, R.Vanitha, Vishwa R, S.Shalini
Volume:
10
Issue:
2
Grenze ID:
01.GIJET.10.2.716
Pages:
5902-5909
Abstract
Cricket, a highly captivating sport, particularly popular in South Asian countries,
heavily relies on umpires to make critical decisions in accordance with the cricket laws. Given
the fallibility of human judgment, errors occasionally occur, leading to time-consuming reviews
by the third umpire to ensure precision. To address this challenge, We introduce an AI-based
umpiring system aimed at refining decision-making precision and fairness in cricket matches.
Through the integration of computer vision and machine learning, our system scrutinizes live
video feeds to provide umpires with instantaneous decision-making support. Critical events like
ball trajectory and player movements are identified by computer vision algorithms, offering
umpires immediate assistance. The system's capabilities evolve through continuous machine
learning, leveraging historical match data to enhance accuracy progressively. A key feature
includes instant replay provisions, empowering teams to contest decisions and fostering
transparency.Our study underscores the transformative role of AI in sports officiating,
presenting a blueprint for its adoption across various sporting domains.