In this talk, we present work on motion analysis with the goal to automatically extract objects of a video scene by using color segmentation and an affine motion model. We give an overview of the whole system and discuss in more detail some key steps of the algorithm such as the clustering of motion vectors and optimization procedure. The segmented video objects can be used afterwards for the generation of mixed-reality scenarios.
M. Gelautz, M. Bleyer, C. Rhemann: "Motion Analysis for Video Object Segmentation"; Talk: Sitzung der ITG-Fachgruppe Algorithmen für die Signalverarbeitung "MIMO- und mehrdimensionale Signalverarbeitung", Wien (invited); 10-03-2008.
Click into the text area and press Ctrl+A/Ctrl+C or ⌘+A/⌘+C to copy the BibTeX into your clipboard… or download the BibTeX.