Divide-and-conquer strategies for estimating multiple transparent motions

Cicero Mota; Ingo Stuke; Til Aach; Erhardt Barth

doi:10.1007/978-3-540-69866-1_6

Divide-and-conquer strategies for estimating multiple transparent motions

Cicero Mota, Ingo Stuke, Til Aach, Erhardt Barth

Abstract

Transparent motions are additive or multiplicative superpositions of moving patterns and occur due to reflections, semi-transparencies, and partial occlusions. The estimation of transparent motions remained a challenging nonlinear problem. We here first linearize the problem in a way which makes it accessible to the known methods used for the estimation of single motions, such as structure tensor, regularization, block matching, Fourier methods, etc. We present the results for two motion layers but there is no limit to the number of layers. Finally, we present a way to categorize different transparent-motion patterns based on the rank of a generalized structure tensor.

Original language	English
Title of host publication	IWCM 2004: Complex Motion
Number of pages	12
Volume	3417 LNCS
Publisher	Springer Verlag
Publication date	01.12.2007
Pages	66-77
ISBN (Print)	978-3-540-69864-7
ISBN (Electronic)	978-3-540-69866-1
DOIs	https://doi.org/10.1007/978-3-540-69866-1_6
Publication status	Published - 01.12.2007
Event	1st International Workshop on Complex Motion - Günzburg, Germany Duration: 12.10.2004 → 14.10.2004 Conference number: 70919

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1007/978-3-540-69866-1_6

Cite this

@inproceedings{932d26eb5fe049708aeeefbdf4ef3039,

title = "Divide-and-conquer strategies for estimating multiple transparent motions",

abstract = "Transparent motions are additive or multiplicative superpositions of moving patterns and occur due to reflections, semi-transparencies, and partial occlusions. The estimation of transparent motions remained a challenging nonlinear problem. We here first linearize the problem in a way which makes it accessible to the known methods used for the estimation of single motions, such as structure tensor, regularization, block matching, Fourier methods, etc. We present the results for two motion layers but there is no limit to the number of layers. Finally, we present a way to categorize different transparent-motion patterns based on the rank of a generalized structure tensor.",

author = "Cicero Mota and Ingo Stuke and Til Aach and Erhardt Barth",

year = "2007",

month = dec,

day = "1",

doi = "10.1007/978-3-540-69866-1_6",

language = "English",

isbn = "978-3-540-69864-7",

volume = "3417 LNCS",

series = "Lecture Notes in Computer Science",

publisher = "Springer Verlag",

pages = " 66--77",

booktitle = "IWCM 2004: Complex Motion",

note = "1st International Workshop on Complex Motion<br/>, IWCM 2004 ; Conference date: 12-10-2004 Through 14-10-2004",

}

Mota, C, Stuke, I, Aach, T & Barth, E 2007, Divide-and-conquer strategies for estimating multiple transparent motions. in IWCM 2004: Complex Motion . vol. 3417 LNCS, Lecture Notes in Computer Science, vol. 3417 LNCS, Springer Verlag, pp. 66-77, 1st International Workshop on Complex Motion
, Günzburg, Bavaria, Germany, 12.10.04. https://doi.org/10.1007/978-3-540-69866-1_6

TY - GEN

T1 - Divide-and-conquer strategies for estimating multiple transparent motions

AU - Mota, Cicero

AU - Stuke, Ingo

AU - Aach, Til

AU - Barth, Erhardt

N1 - Conference code: 70919

PY - 2007/12/1

Y1 - 2007/12/1

N2 - Transparent motions are additive or multiplicative superpositions of moving patterns and occur due to reflections, semi-transparencies, and partial occlusions. The estimation of transparent motions remained a challenging nonlinear problem. We here first linearize the problem in a way which makes it accessible to the known methods used for the estimation of single motions, such as structure tensor, regularization, block matching, Fourier methods, etc. We present the results for two motion layers but there is no limit to the number of layers. Finally, we present a way to categorize different transparent-motion patterns based on the rank of a generalized structure tensor.

AB - Transparent motions are additive or multiplicative superpositions of moving patterns and occur due to reflections, semi-transparencies, and partial occlusions. The estimation of transparent motions remained a challenging nonlinear problem. We here first linearize the problem in a way which makes it accessible to the known methods used for the estimation of single motions, such as structure tensor, regularization, block matching, Fourier methods, etc. We present the results for two motion layers but there is no limit to the number of layers. Finally, we present a way to categorize different transparent-motion patterns based on the rank of a generalized structure tensor.

UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-38049165262&origin=inward&txGid=99a04ee35972cb6f83a3ce80dd8be54b

U2 - 10.1007/978-3-540-69866-1_6

DO - 10.1007/978-3-540-69866-1_6

M3 - Conference contribution

SN - 978-3-540-69864-7

VL - 3417 LNCS

T3 - Lecture Notes in Computer Science

SP - 66

EP - 77

BT - IWCM 2004: Complex Motion

PB - Springer Verlag

T2 - 1st International Workshop on Complex Motion<br/>

Y2 - 12 October 2004 through 14 October 2004

ER -

Divide-and-conquer strategies for estimating multiple transparent motions

Abstract

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this