Accurate background points detection for action recognition in practical video datasets

研究成果: 著書/レポートタイプへの貢献会議での発言

1 引用 (Scopus)

抄録

This paper treats the action recognition of moving objects such as humans in practical video-bases of unconstrained videos. For that, the motion of background image pixels, caused by the camera motion, in video frames of a scene strongly affects the accuracy. Removing such background moving influence is a challenging problem. State-of-The-Art human action recognition approaches adopted human detection to exclude background or directly use some largest motion pattern clusters as candidate background pixels. In this paper, the authors propose a new background estimation method according to background nature. This method could estimate moving background pixels or camera motion from the selected background pixels. Compared with foreground pixels, background pixels have a high diversity and small average distances to the several borders of a video frame. Those are important criteria to estimate background pixels. The proposed approach is based on the long term analysis of point motion trajectories which are more suitable for video image processing. Experimental results show that the proposed approach achieves a very competitive background extraction performance for practical video-bases.

元の言語英語
ホスト出版物のタイトルProceedings of the International Conference on Interfaces and Human Computer Interaction 2016, Game and Entertainment Technologies 2016 and Computer Graphics, Visualization, Computer Vision and Image Processing 2016 - Part of the Multi Conference on Computer Science and Information Systems 2016
出版者IADIS
ページ195-205
ページ数11
ISBN(電子版)9789898533524
出版物ステータス出版済み - 1 1 2016
イベント2016 International Conference on Interfaces and Human Computer Interaction, IHCI 2016, Game and Entertainment Technologies, GET 2016 and Computer Graphics, Visualization, Computer Vision and Image Processing, CGVCVIP 2016 - Madeira, ポルトガル
継続期間: 7 1 20167 4 2016

その他

その他2016 International Conference on Interfaces and Human Computer Interaction, IHCI 2016, Game and Entertainment Technologies, GET 2016 and Computer Graphics, Visualization, Computer Vision and Image Processing, CGVCVIP 2016
ポルトガル
Madeira
期間7/1/167/4/16

Fingerprint

Pixels
Cameras
Image processing
Trajectories

All Science Journal Classification (ASJC) codes

  • Computer Science(all)

これを引用

Xiang, Y., Okada, Y., & Kaneko, K. (2016). Accurate background points detection for action recognition in practical video datasets. : Proceedings of the International Conference on Interfaces and Human Computer Interaction 2016, Game and Entertainment Technologies 2016 and Computer Graphics, Visualization, Computer Vision and Image Processing 2016 - Part of the Multi Conference on Computer Science and Information Systems 2016 (pp. 195-205). IADIS.

Accurate background points detection for action recognition in practical video datasets. / Xiang, Yu; Okada, Yoshihiro; Kaneko, Kosuke.

Proceedings of the International Conference on Interfaces and Human Computer Interaction 2016, Game and Entertainment Technologies 2016 and Computer Graphics, Visualization, Computer Vision and Image Processing 2016 - Part of the Multi Conference on Computer Science and Information Systems 2016. IADIS, 2016. p. 195-205.

研究成果: 著書/レポートタイプへの貢献会議での発言

Xiang, Y, Okada, Y & Kaneko, K 2016, Accurate background points detection for action recognition in practical video datasets. : Proceedings of the International Conference on Interfaces and Human Computer Interaction 2016, Game and Entertainment Technologies 2016 and Computer Graphics, Visualization, Computer Vision and Image Processing 2016 - Part of the Multi Conference on Computer Science and Information Systems 2016. IADIS, pp. 195-205, 2016 International Conference on Interfaces and Human Computer Interaction, IHCI 2016, Game and Entertainment Technologies, GET 2016 and Computer Graphics, Visualization, Computer Vision and Image Processing, CGVCVIP 2016, Madeira, ポルトガル, 7/1/16.
Xiang Y, Okada Y, Kaneko K. Accurate background points detection for action recognition in practical video datasets. : Proceedings of the International Conference on Interfaces and Human Computer Interaction 2016, Game and Entertainment Technologies 2016 and Computer Graphics, Visualization, Computer Vision and Image Processing 2016 - Part of the Multi Conference on Computer Science and Information Systems 2016. IADIS. 2016. p. 195-205
Xiang, Yu ; Okada, Yoshihiro ; Kaneko, Kosuke. / Accurate background points detection for action recognition in practical video datasets. Proceedings of the International Conference on Interfaces and Human Computer Interaction 2016, Game and Entertainment Technologies 2016 and Computer Graphics, Visualization, Computer Vision and Image Processing 2016 - Part of the Multi Conference on Computer Science and Information Systems 2016. IADIS, 2016. pp. 195-205
@inproceedings{2e080caa4c4f4a69b671bf1bb71a381d,
title = "Accurate background points detection for action recognition in practical video datasets",
abstract = "This paper treats the action recognition of moving objects such as humans in practical video-bases of unconstrained videos. For that, the motion of background image pixels, caused by the camera motion, in video frames of a scene strongly affects the accuracy. Removing such background moving influence is a challenging problem. State-of-The-Art human action recognition approaches adopted human detection to exclude background or directly use some largest motion pattern clusters as candidate background pixels. In this paper, the authors propose a new background estimation method according to background nature. This method could estimate moving background pixels or camera motion from the selected background pixels. Compared with foreground pixels, background pixels have a high diversity and small average distances to the several borders of a video frame. Those are important criteria to estimate background pixels. The proposed approach is based on the long term analysis of point motion trajectories which are more suitable for video image processing. Experimental results show that the proposed approach achieves a very competitive background extraction performance for practical video-bases.",
author = "Yu Xiang and Yoshihiro Okada and Kosuke Kaneko",
year = "2016",
month = "1",
day = "1",
language = "English",
pages = "195--205",
booktitle = "Proceedings of the International Conference on Interfaces and Human Computer Interaction 2016, Game and Entertainment Technologies 2016 and Computer Graphics, Visualization, Computer Vision and Image Processing 2016 - Part of the Multi Conference on Computer Science and Information Systems 2016",
publisher = "IADIS",

}

TY - GEN

T1 - Accurate background points detection for action recognition in practical video datasets

AU - Xiang, Yu

AU - Okada, Yoshihiro

AU - Kaneko, Kosuke

PY - 2016/1/1

Y1 - 2016/1/1

N2 - This paper treats the action recognition of moving objects such as humans in practical video-bases of unconstrained videos. For that, the motion of background image pixels, caused by the camera motion, in video frames of a scene strongly affects the accuracy. Removing such background moving influence is a challenging problem. State-of-The-Art human action recognition approaches adopted human detection to exclude background or directly use some largest motion pattern clusters as candidate background pixels. In this paper, the authors propose a new background estimation method according to background nature. This method could estimate moving background pixels or camera motion from the selected background pixels. Compared with foreground pixels, background pixels have a high diversity and small average distances to the several borders of a video frame. Those are important criteria to estimate background pixels. The proposed approach is based on the long term analysis of point motion trajectories which are more suitable for video image processing. Experimental results show that the proposed approach achieves a very competitive background extraction performance for practical video-bases.

AB - This paper treats the action recognition of moving objects such as humans in practical video-bases of unconstrained videos. For that, the motion of background image pixels, caused by the camera motion, in video frames of a scene strongly affects the accuracy. Removing such background moving influence is a challenging problem. State-of-The-Art human action recognition approaches adopted human detection to exclude background or directly use some largest motion pattern clusters as candidate background pixels. In this paper, the authors propose a new background estimation method according to background nature. This method could estimate moving background pixels or camera motion from the selected background pixels. Compared with foreground pixels, background pixels have a high diversity and small average distances to the several borders of a video frame. Those are important criteria to estimate background pixels. The proposed approach is based on the long term analysis of point motion trajectories which are more suitable for video image processing. Experimental results show that the proposed approach achieves a very competitive background extraction performance for practical video-bases.

UR - http://www.scopus.com/inward/record.url?scp=85019216485&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85019216485&partnerID=8YFLogxK

M3 - Conference contribution

SP - 195

EP - 205

BT - Proceedings of the International Conference on Interfaces and Human Computer Interaction 2016, Game and Entertainment Technologies 2016 and Computer Graphics, Visualization, Computer Vision and Image Processing 2016 - Part of the Multi Conference on Computer Science and Information Systems 2016

PB - IADIS

ER -