A Review of Object Segmentation Techniques for Basketball Statistics Estimation

Abstract

With the increased availability of data in basketball games, manual statistics handling has grown increasingly difficult. As a result, automated ways for extracting relevant information from basketball video footage are required.Segmentation, which includes splitting an image into meaningful segments or sections, is an important feature of image analysis. picture segmentation is the process of breaking a picture into meaningful segments or areas, which aids in the identification and analysis of items in the image. Its uses include object detection, medical picture analysis, and computer vision. Despite the potential benefits of correct picture segmentation, it may be a time-consuming and difficult operation, particularly when working with huge datasets. To tackle this difficulty, researchers are attempting to develop automated methods for measuring and estimating picture segmentation outcomes. To extract and analyse characteristics from visual input, these technologies employ a number of methodologies, including machine learning tactics and deep neural networks. In this research paper, we explore techniques that utilize digital image processing and deep learning models to create a referee system that can be automated for basketball games.

Keywords: Segmentation, basketball, picture segmentation, statistics handling