Block grouping algorithm for motion description encoding

S Tredwell; Adrian N Evans

doi:10.1109/ISCAS.2001.922022

Block grouping algorithm for motion description encoding

S Tredwell, Adrian N Evans

Department of Electronic & Electrical Engineering

Research output: Contribution to conference › Paper › peer-review

2 Citations (SciVal)

Abstract

We present a method for improving the overall quality of motion compensated video codecs by using a variable number of motion objects to alter the bandwidth allocated to the motion description. The proposed algorithm takes as its input an existing motion vector field from which it extracts groups of vectors. Control is obtained by only encoding the largest N of these groups. Those blocks which do not belong to the N groups are reassigned to one based upon a block comparison metric. The resulting list of vectors and a mapping showing group membership are then entropy encoded. Experimental results demonstrate that by choosing a value of N that matches the type and complexity of motion in the sequence, the overall compression performance is improved

Original language	English
Pages	V:211-V:214
DOIs	https://doi.org/10.1109/ISCAS.2001.922022
Publication status	Published - 2001
Event	IEEE International Symposium on Circuits and Systems - Sydney, NSW, Australia Duration: 6 May 2001 → 9 May 2001

Conference

Conference	IEEE International Symposium on Circuits and Systems
Country/Territory	Australia
City	Sydney, NSW
Period	6/05/01 → 9/05/01

Access to Document

10.1109/ISCAS.2001.922022

Cite this

@conference{f203fdf8a5b0411d91d71ad4e3fdda71,

title = "Block grouping algorithm for motion description encoding",

abstract = "We present a method for improving the overall quality of motion compensated video codecs by using a variable number of motion objects to alter the bandwidth allocated to the motion description. The proposed algorithm takes as its input an existing motion vector field from which it extracts groups of vectors. Control is obtained by only encoding the largest N of these groups. Those blocks which do not belong to the N groups are reassigned to one based upon a block comparison metric. The resulting list of vectors and a mapping showing group membership are then entropy encoded. Experimental results demonstrate that by choosing a value of N that matches the type and complexity of motion in the sequence, the overall compression performance is improved",

author = "S Tredwell and Evans, {Adrian N}",

year = "2001",

doi = "10.1109/ISCAS.2001.922022",

language = "English",

pages = "V:211--V:214",

note = "IEEE International Symposium on Circuits and Systems ; Conference date: 06-05-2001 Through 09-05-2001",

}

TY - CONF

T1 - Block grouping algorithm for motion description encoding

AU - Tredwell, S

AU - Evans, Adrian N

PY - 2001

Y1 - 2001

N2 - We present a method for improving the overall quality of motion compensated video codecs by using a variable number of motion objects to alter the bandwidth allocated to the motion description. The proposed algorithm takes as its input an existing motion vector field from which it extracts groups of vectors. Control is obtained by only encoding the largest N of these groups. Those blocks which do not belong to the N groups are reassigned to one based upon a block comparison metric. The resulting list of vectors and a mapping showing group membership are then entropy encoded. Experimental results demonstrate that by choosing a value of N that matches the type and complexity of motion in the sequence, the overall compression performance is improved

AB - We present a method for improving the overall quality of motion compensated video codecs by using a variable number of motion objects to alter the bandwidth allocated to the motion description. The proposed algorithm takes as its input an existing motion vector field from which it extracts groups of vectors. Control is obtained by only encoding the largest N of these groups. Those blocks which do not belong to the N groups are reassigned to one based upon a block comparison metric. The resulting list of vectors and a mapping showing group membership are then entropy encoded. Experimental results demonstrate that by choosing a value of N that matches the type and complexity of motion in the sequence, the overall compression performance is improved

UR - http://dx.doi.org/10.1109/ISCAS.2001.922022

U2 - 10.1109/ISCAS.2001.922022

DO - 10.1109/ISCAS.2001.922022

M3 - Paper

SP - V:211-V:214

T2 - IEEE International Symposium on Circuits and Systems

Y2 - 6 May 2001 through 9 May 2001

ER -

Block grouping algorithm for motion description encoding

Abstract

Conference

Access to Document

Other files and links

Fingerprint

Cite this