Dynamic visual tracking using SDG-like matching

Research output: Contribution to journalArticle

Abstract

Motion detection/estimation plays a crucial role in dynamic visual tracking. Whether a dynamic visual tracking system can successfully track a moving target closely depends on the quality of motion detection/estimation results. In dynamic visual tracking, the camera used to capture images is not stationary, so any slight vibration of the camera motion or the target motion can lead to a blurry image causing the visual tracking performance to be deteriorated. To cope with this difficulty, a motion detection/estimation approach consisting of a region-based spatial distribution of Gaussians (SDG)-like matching algorithm, a template-update-with-memory algorithm, and a template mask is developed in this study. Moreover, linear interpolation on vision commands is performed to improve the tracking performance. A dynamic visual tracking system designed for locking the target's image in the center of the image plane is used as the test platform. Experimental results demonstrate the effectiveness of the proposed approach.

Original languageEnglish
Pages (from-to)673-690
Number of pages18
JournalJournal of Information Science and Engineering
Volume24
Issue number3
Publication statusPublished - 2008 May 1

Fingerprint

Spatial distribution
Cameras
performance
Masks
Interpolation
Data storage equipment

All Science Journal Classification (ASJC) codes

  • Software
  • Human-Computer Interaction
  • Hardware and Architecture
  • Library and Information Sciences
  • Computational Theory and Mathematics

Cite this

@article{4cba3f5ee83d47b987857db59fc5fda1,
title = "Dynamic visual tracking using SDG-like matching",
abstract = "Motion detection/estimation plays a crucial role in dynamic visual tracking. Whether a dynamic visual tracking system can successfully track a moving target closely depends on the quality of motion detection/estimation results. In dynamic visual tracking, the camera used to capture images is not stationary, so any slight vibration of the camera motion or the target motion can lead to a blurry image causing the visual tracking performance to be deteriorated. To cope with this difficulty, a motion detection/estimation approach consisting of a region-based spatial distribution of Gaussians (SDG)-like matching algorithm, a template-update-with-memory algorithm, and a template mask is developed in this study. Moreover, linear interpolation on vision commands is performed to improve the tracking performance. A dynamic visual tracking system designed for locking the target's image in the center of the image plane is used as the test platform. Experimental results demonstrate the effectiveness of the proposed approach.",
author = "Ming-Yang Cheng and Mi-Ching Tsai and Chen, {Chun Jen}",
year = "2008",
month = "5",
day = "1",
language = "English",
volume = "24",
pages = "673--690",
journal = "Journal of Information Science and Engineering",
issn = "1016-2364",
publisher = "Institute of Information Science",
number = "3",

}

Dynamic visual tracking using SDG-like matching. / Cheng, Ming-Yang; Tsai, Mi-Ching; Chen, Chun Jen.

In: Journal of Information Science and Engineering, Vol. 24, No. 3, 01.05.2008, p. 673-690.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Dynamic visual tracking using SDG-like matching

AU - Cheng, Ming-Yang

AU - Tsai, Mi-Ching

AU - Chen, Chun Jen

PY - 2008/5/1

Y1 - 2008/5/1

N2 - Motion detection/estimation plays a crucial role in dynamic visual tracking. Whether a dynamic visual tracking system can successfully track a moving target closely depends on the quality of motion detection/estimation results. In dynamic visual tracking, the camera used to capture images is not stationary, so any slight vibration of the camera motion or the target motion can lead to a blurry image causing the visual tracking performance to be deteriorated. To cope with this difficulty, a motion detection/estimation approach consisting of a region-based spatial distribution of Gaussians (SDG)-like matching algorithm, a template-update-with-memory algorithm, and a template mask is developed in this study. Moreover, linear interpolation on vision commands is performed to improve the tracking performance. A dynamic visual tracking system designed for locking the target's image in the center of the image plane is used as the test platform. Experimental results demonstrate the effectiveness of the proposed approach.

AB - Motion detection/estimation plays a crucial role in dynamic visual tracking. Whether a dynamic visual tracking system can successfully track a moving target closely depends on the quality of motion detection/estimation results. In dynamic visual tracking, the camera used to capture images is not stationary, so any slight vibration of the camera motion or the target motion can lead to a blurry image causing the visual tracking performance to be deteriorated. To cope with this difficulty, a motion detection/estimation approach consisting of a region-based spatial distribution of Gaussians (SDG)-like matching algorithm, a template-update-with-memory algorithm, and a template mask is developed in this study. Moreover, linear interpolation on vision commands is performed to improve the tracking performance. A dynamic visual tracking system designed for locking the target's image in the center of the image plane is used as the test platform. Experimental results demonstrate the effectiveness of the proposed approach.

UR - http://www.scopus.com/inward/record.url?scp=44649145854&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=44649145854&partnerID=8YFLogxK

M3 - Article

VL - 24

SP - 673

EP - 690

JO - Journal of Information Science and Engineering

JF - Journal of Information Science and Engineering

SN - 1016-2364

IS - 3

ER -