Multi-view analysis of sports video

Font Standard Bold
Login

Research unit

INNOSUISSE

Project number

9637.1;6 PFES-ES

Project title

Multi-view analysis of sports video

Texts for this project

	German	French	Italian	English
Short description	-	-	-
Abstract	-	-	-

Inserted texts

Category	Text
Short description (English)	Multi-view analysis of sports video
Abstract (English)	In this CTI Project, ETH Zurich and LiberoVision AG will investigate and design novel methods for the analysis of sports video feeds. The interpretation of raw video is one of the main challenges of Computer Vision. The products of LiberoVision will have a great benefit from automatically detecting the contents of the video feeds. The research in this project aims to find novel methods for real-time analysis of the feeds from multiple videos from the same scene and derive an abstract representation of the contents useable for further processing. Novel data-structures and algorithms are required to achieve this goal. These will enable Libero- Vision to keep its technological lead in virtual replays of sports broadcasts.

Category

Text

Short description
(English)

Multi-view analysis of sports video

Abstract
(English)

In this CTI Project, ETH Zurich and LiberoVision AG will investigate and design novel methods for the analysis of sports video feeds. The interpretation of raw video is one of the main challenges of Computer Vision. The products of LiberoVision will have a great benefit from automatically detecting the contents of the video feeds. The research in this project aims to find novel methods for real-time analysis of the feeds from multiple videos from the same scene and derive an abstract representation of the contents useable for further processing. Novel data-structures and algorithms are required to achieve this goal. These will enable Libero- Vision to keep its technological lead in virtual replays of sports broadcasts.