Title :
Joint global motion estimation and coding for scalable H.264/SVC high-definition video streams
Author :
Kas, Christian ; Nicolas, Henri
Author_Institution :
LaBRI UMR CNRS, Univ. de Bordeaux Domaine, Talence
Abstract :
This paper presents a joint indexing-coding approach applied to global camera motion detection in the scalable H.264/SVC compressed domain. Our goal is to facilitate and to improve indexing in the MPEG compressed domain, if necessary by modifying the original coded stream, without losing compatibility with the standard. In frames with very noisy motion vector fields, we use the global motion information obtained by analyzing the raw video to modify the vectors of chosen macroblocks. This results in slightly decreased coding efficiency, but improved and more reliable indexing results in the compressed domain.
Keywords :
data compression; motion estimation; video coding; video streaming; MPEG compressed domain; camera motion detection; compressed domain; global motion estimation; high-definition video streams; indexing-coding approach; motion vector fields; scalable H.264/SVC; Cameras; High definition video; Indexing; Motion analysis; Motion detection; Motion estimation; Static VAr compensators; Streaming media; Transform coding; Video compression; Compressed Domain; Global Motion Estimation; H.264/SVC; Indexing; Scalability;
Conference_Titel :
Content-Based Multimedia Indexing, 2008. CBMI 2008. International Workshop on
Conference_Location :
London
Print_ISBN :
978-1-4244-2043-8
Electronic_ISBN :
978-1-4244-2044-5
DOI :
10.1109/CBMI.2008.4564979