Vehicle-Type Motion Estimation from Multi-frame Images

Yuncai Liu; Thomas S. Huang

doi:10.1109/34.236249

Vehicle-Type Motion Estimation from Multi-frame Images

Yuncai Liu, Thomas S. Huang

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

Abstract

A new model for vehicle-type motion, which assumes that the motion is a rotation around an axis through the vehicle center followed by a forward translation along the main axis of the vehicle, is proposed. The contribution of this paper is threefold: 1) When the rotation and the amplitude of translation are constant, this type of motion is shown to be equivalent to a constant camera-centered motion. This indicates that a constant motion in the conventional camera-centered model, which is commonly considered to be artificial, can in fact be a reasonable model in real life. 2) We show that a constant vehicle-type motion can be interpreted as a constant screw motion. 3) A linear algorithm for estimating constant vehicle-type motion is presented, and experiments using real scene images are included. As an extension, vehicle-type motion with constant rotation and constantly accelerated translation is also discussed.

Original language	English (US)
Pages (from-to)	802-808
Number of pages	7
Journal	IEEE transactions on pattern analysis and machine intelligence
Volume	15
Issue number	8
DOIs	https://doi.org/10.1109/34.236249
State	Published - Aug 1993
Externally published	Yes

Keywords

Constant motion
long image sequence
motion model
point correspondences
vehicle motion

ASJC Scopus subject areas

Software
Computer Vision and Pattern Recognition
Computational Theory and Mathematics
Artificial Intelligence
Applied Mathematics

Online availability

10.1109/34.236249

Library availability

Discover UIUC Full Text

Cite this

@article{924f415169604af0b232ea86871e661d,

title = "Vehicle-Type Motion Estimation from Multi-frame Images",

abstract = "A new model for vehicle-type motion, which assumes that the motion is a rotation around an axis through the vehicle center followed by a forward translation along the main axis of the vehicle, is proposed. The contribution of this paper is threefold: 1) When the rotation and the amplitude of translation are constant, this type of motion is shown to be equivalent to a constant camera-centered motion. This indicates that a constant motion in the conventional camera-centered model, which is commonly considered to be artificial, can in fact be a reasonable model in real life. 2) We show that a constant vehicle-type motion can be interpreted as a constant screw motion. 3) A linear algorithm for estimating constant vehicle-type motion is presented, and experiments using real scene images are included. As an extension, vehicle-type motion with constant rotation and constantly accelerated translation is also discussed.",

keywords = "Constant motion, long image sequence, motion model, point correspondences, vehicle motion",

author = "Yuncai Liu and Huang, {Thomas S.}",

note = "Funding Information: In this correspondence, a new model for vehicle-type motion is proposed. This model represents a 3-D motion as a rotation around an axis through the vehicle center followed by a forward translation. Several advantages arise with this model: First, this model is much closer to real vehicle motion than existing motion models, especially when the motion is small. Second, in the case of constant motion for which the rotation and amplitude of translation are assumed invariant, as we shall see later, a vehicle-type motion is equivalent to a constant camera-centered motion. This gives rise to several benefits: 1) Constant camera-centered motion, which has been used by many researchers, is commonly thought to be artificial, is., few cases of object motion in real life could fit the assumption. The research of this correspondence, however, discovers that constant camera-centered motion is, in fact, a good model for vehicle motion. 2) Since a constant vehicle-type motion can be converted into a constant Manuscript received April 4, 1991; revised October 12, 1991. This work was supported by the National Science Foundation under Grant IRI-89-02 728 and the State of Illinois Department of Commerce and Community Affairs under Grant SCCA 90-82 140. Recommended for acceptance by Associate Editor Y. Aloimonos.",

year = "1993",

month = aug,

doi = "10.1109/34.236249",

language = "English (US)",

volume = "15",

pages = "802--808",

journal = "IEEE transactions on pattern analysis and machine intelligence",

issn = "0162-8828",

publisher = "IEEE Computer Society",

number = "8",

}

TY - JOUR

T1 - Vehicle-Type Motion Estimation from Multi-frame Images

AU - Liu, Yuncai

AU - Huang, Thomas S.

N1 - Funding Information: In this correspondence, a new model for vehicle-type motion is proposed. This model represents a 3-D motion as a rotation around an axis through the vehicle center followed by a forward translation. Several advantages arise with this model: First, this model is much closer to real vehicle motion than existing motion models, especially when the motion is small. Second, in the case of constant motion for which the rotation and amplitude of translation are assumed invariant, as we shall see later, a vehicle-type motion is equivalent to a constant camera-centered motion. This gives rise to several benefits: 1) Constant camera-centered motion, which has been used by many researchers, is commonly thought to be artificial, is., few cases of object motion in real life could fit the assumption. The research of this correspondence, however, discovers that constant camera-centered motion is, in fact, a good model for vehicle motion. 2) Since a constant vehicle-type motion can be converted into a constant Manuscript received April 4, 1991; revised October 12, 1991. This work was supported by the National Science Foundation under Grant IRI-89-02 728 and the State of Illinois Department of Commerce and Community Affairs under Grant SCCA 90-82 140. Recommended for acceptance by Associate Editor Y. Aloimonos.

PY - 1993/8

Y1 - 1993/8

N2 - A new model for vehicle-type motion, which assumes that the motion is a rotation around an axis through the vehicle center followed by a forward translation along the main axis of the vehicle, is proposed. The contribution of this paper is threefold: 1) When the rotation and the amplitude of translation are constant, this type of motion is shown to be equivalent to a constant camera-centered motion. This indicates that a constant motion in the conventional camera-centered model, which is commonly considered to be artificial, can in fact be a reasonable model in real life. 2) We show that a constant vehicle-type motion can be interpreted as a constant screw motion. 3) A linear algorithm for estimating constant vehicle-type motion is presented, and experiments using real scene images are included. As an extension, vehicle-type motion with constant rotation and constantly accelerated translation is also discussed.

AB - A new model for vehicle-type motion, which assumes that the motion is a rotation around an axis through the vehicle center followed by a forward translation along the main axis of the vehicle, is proposed. The contribution of this paper is threefold: 1) When the rotation and the amplitude of translation are constant, this type of motion is shown to be equivalent to a constant camera-centered motion. This indicates that a constant motion in the conventional camera-centered model, which is commonly considered to be artificial, can in fact be a reasonable model in real life. 2) We show that a constant vehicle-type motion can be interpreted as a constant screw motion. 3) A linear algorithm for estimating constant vehicle-type motion is presented, and experiments using real scene images are included. As an extension, vehicle-type motion with constant rotation and constantly accelerated translation is also discussed.

KW - Constant motion

KW - long image sequence

KW - motion model

KW - point correspondences

KW - vehicle motion

UR - http://www.scopus.com/inward/record.url?scp=0027644490&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0027644490&partnerID=8YFLogxK

U2 - 10.1109/34.236249

DO - 10.1109/34.236249

M3 - Article

AN - SCOPUS:0027644490

SN - 0162-8828

VL - 15

SP - 802

EP - 808

JO - IEEE transactions on pattern analysis and machine intelligence

JF - IEEE transactions on pattern analysis and machine intelligence

IS - 8

ER -

Vehicle-Type Motion Estimation from Multi-frame Images

Abstract

Keywords

ASJC Scopus subject areas

Online availability

Library availability

Related links

Fingerprint

Cite this