A fast algorithm for detection of camera motion

Hyeokman Kim, Tae Hoon Kwon, Woon Kyung Kim, Byung D. Rhee, S. M H Song

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

1 Citation (Scopus)

Abstract

Most image analysis/understanding applications require accurate computation of camera motion parameters. However, in multimedia applications, particularly in video parsing, the exact camera motion parameters such as the panning and/or zooming rates are not needed. The detection-i.e., a binary decision-of camera motion is all that is required to avoid declaring a false scene change. As camera motions can induce false scene changes for video parsing algorithms, we propose a fast algorithm to detect such camera motions: camera zoom and pan. As the algorithm is only expected produce a binary decision, without the exact panning/zooming rates, the proposed algorithm runs on a reduced data set, namely the projection data. The algorithm begins with a central portion of the image and computes the projection data (or the line integrals along the x- or y- axis) to turn the 2-D image data into a 1-D data. This projected 1-D data is further processed via correlation processing to detect camera zoom and pan. Working with projection data saves processing time by an order of magnitude, since for instance, a 2-D correlation takes N 2 multiplies per point, however a 1-D correlation takes N multiplies per point. The efficacy of the proposed algorithm is tested for a number of image sequences and the algorithm is shown to be successful in detecting camera motions. The proposed algorithm is expected to be beneficial for video parsers working with Motion-JPEG data stream where motion vectors are not available.

Original languageEnglish
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
EditorsD. Sinha
Pages78-87
Number of pages10
Volume3303
DOIs
Publication statusPublished - 1998
Externally publishedYes
EventReal-Time Imaging III - San Jose, CA, United States
Duration: 1998 Jan 261998 Jan 26

Other

OtherReal-Time Imaging III
CountryUnited States
CitySan Jose, CA
Period98/1/2698/1/26

Fingerprint

Cameras
cameras
projection
parsing algorithms
multimedia
image analysis
Image analysis
Processing

Keywords

  • Camera motion detection
  • Scene change detection
  • Video library
  • Video parser

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Condensed Matter Physics

Cite this

Kim, H., Kwon, T. H., Kim, W. K., Rhee, B. D., & Song, S. M. H. (1998). A fast algorithm for detection of camera motion. In D. Sinha (Ed.), Proceedings of SPIE - The International Society for Optical Engineering (Vol. 3303, pp. 78-87) https://doi.org/10.1117/12.302417

A fast algorithm for detection of camera motion. / Kim, Hyeokman; Kwon, Tae Hoon; Kim, Woon Kyung; Rhee, Byung D.; Song, S. M H.

Proceedings of SPIE - The International Society for Optical Engineering. ed. / D. Sinha. Vol. 3303 1998. p. 78-87.

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

Kim, H, Kwon, TH, Kim, WK, Rhee, BD & Song, SMH 1998, A fast algorithm for detection of camera motion. in D Sinha (ed.), Proceedings of SPIE - The International Society for Optical Engineering. vol. 3303, pp. 78-87, Real-Time Imaging III, San Jose, CA, United States, 98/1/26. https://doi.org/10.1117/12.302417
Kim H, Kwon TH, Kim WK, Rhee BD, Song SMH. A fast algorithm for detection of camera motion. In Sinha D, editor, Proceedings of SPIE - The International Society for Optical Engineering. Vol. 3303. 1998. p. 78-87 https://doi.org/10.1117/12.302417
Kim, Hyeokman ; Kwon, Tae Hoon ; Kim, Woon Kyung ; Rhee, Byung D. ; Song, S. M H. / A fast algorithm for detection of camera motion. Proceedings of SPIE - The International Society for Optical Engineering. editor / D. Sinha. Vol. 3303 1998. pp. 78-87
@inproceedings{ac2d9fd315214f4cb36ae8a354e6d187,
title = "A fast algorithm for detection of camera motion",
abstract = "Most image analysis/understanding applications require accurate computation of camera motion parameters. However, in multimedia applications, particularly in video parsing, the exact camera motion parameters such as the panning and/or zooming rates are not needed. The detection-i.e., a binary decision-of camera motion is all that is required to avoid declaring a false scene change. As camera motions can induce false scene changes for video parsing algorithms, we propose a fast algorithm to detect such camera motions: camera zoom and pan. As the algorithm is only expected produce a binary decision, without the exact panning/zooming rates, the proposed algorithm runs on a reduced data set, namely the projection data. The algorithm begins with a central portion of the image and computes the projection data (or the line integrals along the x- or y- axis) to turn the 2-D image data into a 1-D data. This projected 1-D data is further processed via correlation processing to detect camera zoom and pan. Working with projection data saves processing time by an order of magnitude, since for instance, a 2-D correlation takes N 2 multiplies per point, however a 1-D correlation takes N multiplies per point. The efficacy of the proposed algorithm is tested for a number of image sequences and the algorithm is shown to be successful in detecting camera motions. The proposed algorithm is expected to be beneficial for video parsers working with Motion-JPEG data stream where motion vectors are not available.",
keywords = "Camera motion detection, Scene change detection, Video library, Video parser",
author = "Hyeokman Kim and Kwon, {Tae Hoon} and Kim, {Woon Kyung} and Rhee, {Byung D.} and Song, {S. M H}",
year = "1998",
doi = "10.1117/12.302417",
language = "English",
volume = "3303",
pages = "78--87",
editor = "D. Sinha",
booktitle = "Proceedings of SPIE - The International Society for Optical Engineering",

}

TY - GEN

T1 - A fast algorithm for detection of camera motion

AU - Kim, Hyeokman

AU - Kwon, Tae Hoon

AU - Kim, Woon Kyung

AU - Rhee, Byung D.

AU - Song, S. M H

PY - 1998

Y1 - 1998

N2 - Most image analysis/understanding applications require accurate computation of camera motion parameters. However, in multimedia applications, particularly in video parsing, the exact camera motion parameters such as the panning and/or zooming rates are not needed. The detection-i.e., a binary decision-of camera motion is all that is required to avoid declaring a false scene change. As camera motions can induce false scene changes for video parsing algorithms, we propose a fast algorithm to detect such camera motions: camera zoom and pan. As the algorithm is only expected produce a binary decision, without the exact panning/zooming rates, the proposed algorithm runs on a reduced data set, namely the projection data. The algorithm begins with a central portion of the image and computes the projection data (or the line integrals along the x- or y- axis) to turn the 2-D image data into a 1-D data. This projected 1-D data is further processed via correlation processing to detect camera zoom and pan. Working with projection data saves processing time by an order of magnitude, since for instance, a 2-D correlation takes N 2 multiplies per point, however a 1-D correlation takes N multiplies per point. The efficacy of the proposed algorithm is tested for a number of image sequences and the algorithm is shown to be successful in detecting camera motions. The proposed algorithm is expected to be beneficial for video parsers working with Motion-JPEG data stream where motion vectors are not available.

AB - Most image analysis/understanding applications require accurate computation of camera motion parameters. However, in multimedia applications, particularly in video parsing, the exact camera motion parameters such as the panning and/or zooming rates are not needed. The detection-i.e., a binary decision-of camera motion is all that is required to avoid declaring a false scene change. As camera motions can induce false scene changes for video parsing algorithms, we propose a fast algorithm to detect such camera motions: camera zoom and pan. As the algorithm is only expected produce a binary decision, without the exact panning/zooming rates, the proposed algorithm runs on a reduced data set, namely the projection data. The algorithm begins with a central portion of the image and computes the projection data (or the line integrals along the x- or y- axis) to turn the 2-D image data into a 1-D data. This projected 1-D data is further processed via correlation processing to detect camera zoom and pan. Working with projection data saves processing time by an order of magnitude, since for instance, a 2-D correlation takes N 2 multiplies per point, however a 1-D correlation takes N multiplies per point. The efficacy of the proposed algorithm is tested for a number of image sequences and the algorithm is shown to be successful in detecting camera motions. The proposed algorithm is expected to be beneficial for video parsers working with Motion-JPEG data stream where motion vectors are not available.

KW - Camera motion detection

KW - Scene change detection

KW - Video library

KW - Video parser

UR - http://www.scopus.com/inward/record.url?scp=0032374234&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032374234&partnerID=8YFLogxK

U2 - 10.1117/12.302417

DO - 10.1117/12.302417

M3 - Conference contribution

AN - SCOPUS:0032374234

VL - 3303

SP - 78

EP - 87

BT - Proceedings of SPIE - The International Society for Optical Engineering

A2 - Sinha, D.

ER -