Improved image segmentation using motion

L. Howell; P. Child; P. Hall

doi:10.1145/2534008.2534012

Improved image segmentation using motion

L. Howell
, P. Child
, P. Hall

Department of Computer Science

Aardman Animations Ltd.

Research output: Chapter or section in a book/report/conference proceeding › Book chapter

Abstract

Existing video segmentation methods struggle to work accurately on production images. In particular, methods that rely on colour analysis can suffer when object and background colour ranges are similar. We extend one state of the art method by incorporating motion information from video frames. When tested on a variety of production footage, our new method shows significantly improved results.

Original language	English
Title of host publication	ACM International Conference Proceeding Series
DOIs	https://doi.org/10.1145/2534008.2534012
Publication status	Published - 1 Jan 2013

Access to Document

10.1145/2534008.2534012

Cite this

@inbook{e8cfa34533ec4c7790f0922b84e6c234,

title = "Improved image segmentation using motion",

abstract = "Existing video segmentation methods struggle to work accurately on production images. In particular, methods that rely on colour analysis can suffer when object and background colour ranges are similar. We extend one state of the art method by incorporating motion information from video frames. When tested on a variety of production footage, our new method shows significantly improved results.",

author = "L. Howell and P. Child and P. Hall",

year = "2013",

month = jan,

day = "1",

doi = "10.1145/2534008.2534012",

language = "English",

isbn = "9781450325899",

booktitle = "ACM International Conference Proceeding Series",

}

TY - CHAP

T1 - Improved image segmentation using motion

AU - Howell, L.

AU - Child, P.

AU - Hall, P.

PY - 2013/1/1

Y1 - 2013/1/1

N2 - Existing video segmentation methods struggle to work accurately on production images. In particular, methods that rely on colour analysis can suffer when object and background colour ranges are similar. We extend one state of the art method by incorporating motion information from video frames. When tested on a variety of production footage, our new method shows significantly improved results.

AB - Existing video segmentation methods struggle to work accurately on production images. In particular, methods that rely on colour analysis can suffer when object and background colour ranges are similar. We extend one state of the art method by incorporating motion information from video frames. When tested on a variety of production footage, our new method shows significantly improved results.

UR - https://www.scopus.com/pages/publications/84897598423

UR - http://dx.doi.org/10.1145/2534008.2534012

U2 - 10.1145/2534008.2534012

DO - 10.1145/2534008.2534012

M3 - Book chapter

AN - SCOPUS:84897598423

SN - 9781450325899

BT - ACM International Conference Proceeding Series

ER -

Improved image segmentation using motion

Abstract

Access to Document

Other files and links

Fingerprint

Cite this