Human Head Tracking for Video Indexing

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F07%3A03144361" target="_blank" >RIV/68407700:21230/07:03144361 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Human Head Tracking for Video Indexing

Popis výsledku v původním jazyce

The objective of this research is to design and experimentally evaluate algorithms for human head tracking using a single camera. We summarize the existing approaches on 2D, rigid 3D and nonrigid 3D head tracking and their applications in automatic videoindexing. We compare several different search strategies and criterial functions for tracking with a textured ellipsoidal head model and Ransac-based approach with a point-based model and bundle adjustment. We also describe and evaluate 2 new approachesfor combination of contour and point features. We outline the plans for future work. We plan to extend and evaluate the robustness of the methods, explore the other possibilities of combining different features and deal with the problem of resource allocation for different algorithms used in video indexing.

Název v anglickém jazyce

Human Head Tracking for Video Indexing

Popis výsledku anglicky

The objective of this research is to design and experimentally evaluate algorithms for human head tracking using a single camera. We summarize the existing approaches on 2D, rigid 3D and nonrigid 3D head tracking and their applications in automatic videoindexing. We compare several different search strategies and criterial functions for tracking with a textured ellipsoidal head model and Ransac-based approach with a point-based model and bundle adjustment. We also describe and evaluate 2 new approachesfor combination of contour and point features. We outline the plans for future work. We plan to extend and evaluate the robustness of the methods, explore the other possibilities of combining different features and deal with the problem of resource allocation for different algorithms used in video indexing.

Druh

O - Ostatní výsledky

CEP obor

JD - Využití počítačů, robotika a její aplikace

OECD FORD obor

—

Projekt

<a href="/cs/project/1ET101210407" target="_blank" >1ET101210407: Multikamerový systém pro modelování a rozpoznávání událostí</a><br>

Návaznosti

V - Vyzkumna aktivita podporovana z jinych verejnych zdroju

Rok uplatnění

2007

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů