Title
Characterization of Slice-Based H.264/AVC Encoded Video Traffic
Abstract
An explicit slice-based mode type selection scheme for use in H.264/AVC has recently been developed, which preserves the coding efficiency of H.264/AVC without requiring any modifications to the standard. The novel scheme avoids the burstiness effect of standard frame-based H.264/AVC by breaking up the Group Of Pictures (GOP) structure and each frame contains an intracoded slice at successive positions. A large intracoded frame is then only present at the beginning of each sequence. In addition, the first frame in each scene is large because of the prediction error when trying to predict the current frame based on the last frame in the previous scene. In this work a single slice-based encoded video stream is studied. The characteristics of the video stream divided into scenes using a scene change detection algorithm are investigated, looking at the marginal distributions and the correlations. A simulation study, looking at the multiplexing properties and the buffer overflow probabilities for a slice-based encoded stream compared to a frame-based encoded stream is also performed. The study shows that frame correlation is present only at small lags, and the periodic correlation structure seen in frame-based encoded video is mostly removed for the stream under study. Results from simulations show that the slice-based encoded video stream performs better than the frame-based encoded video stream in terms of lower packet loss rates and lower packet delay as long as the buffer size is small or the link utilization is moderate.
Year
DOI
Venue
2007
10.1109/ECUMN.2007.10
ECUMN
Keywords
Field
DocType
traffic simulations,current frame,slice-based video coding,slice-based h.264,last frame,slice-based encoded video stream,slice-based encoded stream,avc encoded video traffic,large intracoded frame,video stream,video characteriza- tion,frame-based encoded video,frame-based encoded stream,frame correlation,frame-based encoded video stream,prediction error,buffer overflow
Reference frame,Algorithmic efficiency,Group of pictures,Computer science,Network packet,Algorithm,Real-time computing,Residual frame,Burstiness,Frame (networking),Inter frame
Conference
ISBN
Citations 
PageRank 
0-7695-2768-X
5
0.52
References 
Authors
10
3
Name
Order
Citations
PageRank
Astrid Undheim11049.44
Yuan Lin2192.31
Peder J. Emstad3585.85