Multi-view and multi-frame analysis of video footage

• Unité de recherche: INNOSUISSE
• Numéro de projet: 12218.1;8 PFES-ES
• Titre du projet anglais: Multi-view and multi-frame analysis of video footage

Description succincte (Allemand)

Multi-view and multi-frame analysis of video footage

Description succincte (Anglais)

Multi-view and multi-frame analysis of video footage

Résumé des résultats (Abstract) (Allemand)

In this CTI project, ETH Zurich and LiberoVision AG will investigate and design novel methods for free-viewpoint video. The goal is to strengthen LiberoVision's market position through new innovations, while blocking the competitors with new IP. New products taking advantage of a virtual video sequence, or instant replays for multiple sports bare a big potential in the broadcast market. Market research has shown that this extension meets the demands of our customers. The research focus lies on consistent and temporally coherent multi-cam and multi-frame processing. based on the joint optimization of strongly dependent constraints, a robust, efficient and highly automatic method resulting in a free-viewpoint video with minimal artifacts shall be found. The research area lies in multi-camera object cutout/segmentation over time, and multi-camera calibration tracking.

Résumé des résultats (Abstract) (Anglais)

In this CTI project, ETH Zurich and LiberoVision AG will investigate and design novel methods for free-viewpoint video. The goal is to strengthen LiberoVision's market position through new innovations, while blocking the competitors with new IP. New products taking advantage of a virtual video sequence, or instant replays for multiple sports bare a big potential in the broadcast market. Market research has shown that this extension meets the demands of our customers. The research focus lies on consistent and temporally coherent multi-cam and multi-frame processing. based on the joint optimization of strongly dependent constraints, a robust, efficient and highly automatic method resulting in a free-viewpoint video with minimal artifacts shall be found. The research area lies in multi-camera object cutout/segmentation over time, and multi-camera calibration tracking.