Smoothing of optical flow using robustified diffusion kernels

Ashish Doshi; Adrian G. Bors

doi:10.1016/j.imavis.2010.04.001

Smoothing of optical flow using robustified diffusion kernels

Ashish Doshi, Adrian G. Bors

Computer Science

Research output: Contribution to journal › Article › peer-review

Abstract

This paper proposes a new optical flow smoothing methodology combining vector diffusion and robust statistics. Vector smoothing using diffusion preserves moving object boundaries and the main motion discontinuities. According to a study provided in the paper, diffusion does not remove the outliers but spreads them out, introducing a bias in the neighbourhood. In this paper robust statistics operators such as the median and alpha-trimmed mean are considered for robustifying the diffusion kernels. The robust diffusion smoothing process is extended to 3-D lattices as well. The proposed algorithms are applied for smoothing artificially generated vector fields as well as the optical flow estimated from image sequences. (c) 2010 Elsevier B.V. All rights reserved.

Original language	English
Pages (from-to)	1575-1589
Number of pages	15
Journal	Image and Vision Computing
Volume	28
Issue number	12
DOIs	https://doi.org/10.1016/j.imavis.2010.04.001
Publication status	Published - Dec 2010

Keywords

Anisotropic diffusion
Robust statistics
Optical flow smoothing
BASIS FUNCTION NETWORK
ANISOTROPIC DIFFUSION
NONLINEAR DIFFUSION
EDGE-DETECTION
IMAGE
SEGMENTATION
REGULARIZATION
PDES

Access to Document

10.1016/j.imavis.2010.04.001

http://www.sciencedirect.com/science/article/pii/S0262885610000600

Cite this

@article{dc5dc847ec1e4b12abba291aa72d6731,

title = "Smoothing of optical flow using robustified diffusion kernels",

abstract = "This paper proposes a new optical flow smoothing methodology combining vector diffusion and robust statistics. Vector smoothing using diffusion preserves moving object boundaries and the main motion discontinuities. According to a study provided in the paper, diffusion does not remove the outliers but spreads them out, introducing a bias in the neighbourhood. In this paper robust statistics operators such as the median and alpha-trimmed mean are considered for robustifying the diffusion kernels. The robust diffusion smoothing process is extended to 3-D lattices as well. The proposed algorithms are applied for smoothing artificially generated vector fields as well as the optical flow estimated from image sequences. (c) 2010 Elsevier B.V. All rights reserved.",

keywords = "Anisotropic diffusion, Robust statistics, Optical flow smoothing, BASIS FUNCTION NETWORK, ANISOTROPIC DIFFUSION, NONLINEAR DIFFUSION, EDGE-DETECTION, IMAGE, SEGMENTATION, REGULARIZATION, PDES",

author = "Ashish Doshi and Bors, {Adrian G.}",

year = "2010",

month = dec,

doi = "10.1016/j.imavis.2010.04.001",

language = "English",

volume = "28",

pages = "1575--1589",

journal = "Image and Vision Computing",

issn = "0262-8856",

publisher = "Elsevier",

number = "12",

}

TY - JOUR

T1 - Smoothing of optical flow using robustified diffusion kernels

AU - Doshi, Ashish

AU - Bors, Adrian G.

PY - 2010/12

Y1 - 2010/12

N2 - This paper proposes a new optical flow smoothing methodology combining vector diffusion and robust statistics. Vector smoothing using diffusion preserves moving object boundaries and the main motion discontinuities. According to a study provided in the paper, diffusion does not remove the outliers but spreads them out, introducing a bias in the neighbourhood. In this paper robust statistics operators such as the median and alpha-trimmed mean are considered for robustifying the diffusion kernels. The robust diffusion smoothing process is extended to 3-D lattices as well. The proposed algorithms are applied for smoothing artificially generated vector fields as well as the optical flow estimated from image sequences. (c) 2010 Elsevier B.V. All rights reserved.

AB - This paper proposes a new optical flow smoothing methodology combining vector diffusion and robust statistics. Vector smoothing using diffusion preserves moving object boundaries and the main motion discontinuities. According to a study provided in the paper, diffusion does not remove the outliers but spreads them out, introducing a bias in the neighbourhood. In this paper robust statistics operators such as the median and alpha-trimmed mean are considered for robustifying the diffusion kernels. The robust diffusion smoothing process is extended to 3-D lattices as well. The proposed algorithms are applied for smoothing artificially generated vector fields as well as the optical flow estimated from image sequences. (c) 2010 Elsevier B.V. All rights reserved.

KW - Anisotropic diffusion

KW - Robust statistics

KW - Optical flow smoothing

KW - BASIS FUNCTION NETWORK

KW - ANISOTROPIC DIFFUSION

KW - NONLINEAR DIFFUSION

KW - EDGE-DETECTION

KW - IMAGE

KW - SEGMENTATION

KW - REGULARIZATION

KW - PDES

UR - http://www.scopus.com/inward/record.url?scp=77956344041&partnerID=8YFLogxK

U2 - 10.1016/j.imavis.2010.04.001

DO - 10.1016/j.imavis.2010.04.001

M3 - Article

VL - 28

SP - 1575

EP - 1589

JO - Image and Vision Computing

JF - Image and Vision Computing

SN - 0262-8856

IS - 12

ER -

Smoothing of optical flow using robustified diffusion kernels

Abstract

Keywords

Access to Document

Other files and links

Cite this