Low-dimensional superpixel descriptor for visual correspondence estimation in video

Songlin Du, Takeshi Ikenaga

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Estimating local visual correspondence between video frames is an important and essential challenge in many visual applications. Keypoint based sparse matching is a common way to address the problem of local visual correspondence estimation. This paper proposes a local visual correspondence estimation method based on extracting discriminative features from superpixels. In the proposed approach, superpixels are generated by over segmentation at first. Then the superpixels are described by orientated center-boundary distance (OCB-D) and gray-level co-occurrence matrix (GLCM) which extract shape feature and texture feature, respectively. Experimental results on the widely used Middlebury dataset prove that the proposed superpixel descriptor achieves much higher accuracy than the compact ORB descriptor when same dimensions of features are used. In addition, benefited from its low-dimension character, the proposed descriptor is memory-efficient and hardware friendly.

Original languageEnglish
Title of host publication2017 International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2017 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages287-291
Number of pages5
Volume2018-January
ISBN (Electronic)9781538621592
DOIs
Publication statusPublished - 2018 Jan 19
Event25th International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2017 - Xiamen, China
Duration: 2017 Nov 62017 Nov 9

Other

Other25th International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2017
CountryChina
CityXiamen
Period17/11/617/11/9

Fingerprint

Computer hardware
Textures
Data storage equipment

Keywords

  • Local descriptor
  • Low dimension
  • Superpixel
  • Visual correspondence

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Networks and Communications
  • Signal Processing

Cite this

Du, S., & Ikenaga, T. (2018). Low-dimensional superpixel descriptor for visual correspondence estimation in video. In 2017 International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2017 - Proceedings (Vol. 2018-January, pp. 287-291). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ISPACS.2017.8266490

Low-dimensional superpixel descriptor for visual correspondence estimation in video. / Du, Songlin; Ikenaga, Takeshi.

2017 International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2017 - Proceedings. Vol. 2018-January Institute of Electrical and Electronics Engineers Inc., 2018. p. 287-291.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Du, S & Ikenaga, T 2018, Low-dimensional superpixel descriptor for visual correspondence estimation in video. in 2017 International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2017 - Proceedings. vol. 2018-January, Institute of Electrical and Electronics Engineers Inc., pp. 287-291, 25th International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2017, Xiamen, China, 17/11/6. https://doi.org/10.1109/ISPACS.2017.8266490
Du S, Ikenaga T. Low-dimensional superpixel descriptor for visual correspondence estimation in video. In 2017 International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2017 - Proceedings. Vol. 2018-January. Institute of Electrical and Electronics Engineers Inc. 2018. p. 287-291 https://doi.org/10.1109/ISPACS.2017.8266490
Du, Songlin ; Ikenaga, Takeshi. / Low-dimensional superpixel descriptor for visual correspondence estimation in video. 2017 International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2017 - Proceedings. Vol. 2018-January Institute of Electrical and Electronics Engineers Inc., 2018. pp. 287-291
@inproceedings{6b91d6d9be9943f5ab66a7a6e9ae046c,
title = "Low-dimensional superpixel descriptor for visual correspondence estimation in video",
abstract = "Estimating local visual correspondence between video frames is an important and essential challenge in many visual applications. Keypoint based sparse matching is a common way to address the problem of local visual correspondence estimation. This paper proposes a local visual correspondence estimation method based on extracting discriminative features from superpixels. In the proposed approach, superpixels are generated by over segmentation at first. Then the superpixels are described by orientated center-boundary distance (OCB-D) and gray-level co-occurrence matrix (GLCM) which extract shape feature and texture feature, respectively. Experimental results on the widely used Middlebury dataset prove that the proposed superpixel descriptor achieves much higher accuracy than the compact ORB descriptor when same dimensions of features are used. In addition, benefited from its low-dimension character, the proposed descriptor is memory-efficient and hardware friendly.",
keywords = "Local descriptor, Low dimension, Superpixel, Visual correspondence",
author = "Songlin Du and Takeshi Ikenaga",
year = "2018",
month = "1",
day = "19",
doi = "10.1109/ISPACS.2017.8266490",
language = "English",
volume = "2018-January",
pages = "287--291",
booktitle = "2017 International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2017 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Low-dimensional superpixel descriptor for visual correspondence estimation in video

AU - Du, Songlin

AU - Ikenaga, Takeshi

PY - 2018/1/19

Y1 - 2018/1/19

N2 - Estimating local visual correspondence between video frames is an important and essential challenge in many visual applications. Keypoint based sparse matching is a common way to address the problem of local visual correspondence estimation. This paper proposes a local visual correspondence estimation method based on extracting discriminative features from superpixels. In the proposed approach, superpixels are generated by over segmentation at first. Then the superpixels are described by orientated center-boundary distance (OCB-D) and gray-level co-occurrence matrix (GLCM) which extract shape feature and texture feature, respectively. Experimental results on the widely used Middlebury dataset prove that the proposed superpixel descriptor achieves much higher accuracy than the compact ORB descriptor when same dimensions of features are used. In addition, benefited from its low-dimension character, the proposed descriptor is memory-efficient and hardware friendly.

AB - Estimating local visual correspondence between video frames is an important and essential challenge in many visual applications. Keypoint based sparse matching is a common way to address the problem of local visual correspondence estimation. This paper proposes a local visual correspondence estimation method based on extracting discriminative features from superpixels. In the proposed approach, superpixels are generated by over segmentation at first. Then the superpixels are described by orientated center-boundary distance (OCB-D) and gray-level co-occurrence matrix (GLCM) which extract shape feature and texture feature, respectively. Experimental results on the widely used Middlebury dataset prove that the proposed superpixel descriptor achieves much higher accuracy than the compact ORB descriptor when same dimensions of features are used. In addition, benefited from its low-dimension character, the proposed descriptor is memory-efficient and hardware friendly.

KW - Local descriptor

KW - Low dimension

KW - Superpixel

KW - Visual correspondence

UR - http://www.scopus.com/inward/record.url?scp=85047465105&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85047465105&partnerID=8YFLogxK

U2 - 10.1109/ISPACS.2017.8266490

DO - 10.1109/ISPACS.2017.8266490

M3 - Conference contribution

VL - 2018-January

SP - 287

EP - 291

BT - 2017 International Symposium on Intelligent Signal Processing and Communication Systems, ISPACS 2017 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -