Multi-level view synthesis (MLVS) based on depth image layer separation (DILS) algorithm for multi-camera view

Nurulfajar Abd Manap, John Soraghan, Lykourgos Petropoulakis

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

Abstract

A novel Multi-Level View Synthesis (MLVS) approach for 3D vision and free-viewpoint video applications, such as light field imaging, is presented. MLVS exploits the advantages of Depth Image Layer Separation (DILS), a new inter-view interpolation algorithm, by extending stereo to multiple camera configurations. The technique finds the pixel correspondences and synthesis through two levels of matching and synthesis process. The main aim of MLVS is to create a multi-camera view system through a reduced number of actual image acquisition cameras, whilst maintaining the quality of the virtual view synthesis images. The proposed technique is shown to offer improved performance and provide additional views with fewer cameras compared to conventional high volume camera configurations for free-viewpoint video acquisition. Thus, substantial cost savings can ensue in processing, calibration, bandwidth and storage requirements.
LanguageEnglish
Title of host publication2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA)
PublisherIEEE
Pages71-76
Number of pages6
ISBN (Print)9781479902675
DOIs
Publication statusPublished - 10 Oct 2013
Event2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA) - Meloka, Malaysia
Duration: 8 Oct 201310 Oct 2013

Conference

Conference2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA)
CountryMalaysia
CityMeloka
Period8/10/1310/10/13

Fingerprint

Cameras
Image acquisition
Interpolation
Pixels
Calibration
Bandwidth
Imaging techniques
Processing
Costs

Keywords

  • calibration
  • computer vision
  • image matching
  • image resolution
  • image sensors
  • interpolation
  • stereo image processing
  • video signal processing
  • arrays
  • cameras
  • electronics packaging
  • image processing
  • PSNR

Cite this

Abd Manap, N., Soraghan, J., & Petropoulakis, L. (2013). Multi-level view synthesis (MLVS) based on depth image layer separation (DILS) algorithm for multi-camera view. In 2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA) (pp. 71-76). IEEE. https://doi.org/10.1109/ICSIPA.2013.6707980
Abd Manap, Nurulfajar ; Soraghan, John ; Petropoulakis, Lykourgos. / Multi-level view synthesis (MLVS) based on depth image layer separation (DILS) algorithm for multi-camera view. 2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA). IEEE, 2013. pp. 71-76
@inproceedings{fe5824c6e2494517901811b5582108ab,
title = "Multi-level view synthesis (MLVS) based on depth image layer separation (DILS) algorithm for multi-camera view",
abstract = "A novel Multi-Level View Synthesis (MLVS) approach for 3D vision and free-viewpoint video applications, such as light field imaging, is presented. MLVS exploits the advantages of Depth Image Layer Separation (DILS), a new inter-view interpolation algorithm, by extending stereo to multiple camera configurations. The technique finds the pixel correspondences and synthesis through two levels of matching and synthesis process. The main aim of MLVS is to create a multi-camera view system through a reduced number of actual image acquisition cameras, whilst maintaining the quality of the virtual view synthesis images. The proposed technique is shown to offer improved performance and provide additional views with fewer cameras compared to conventional high volume camera configurations for free-viewpoint video acquisition. Thus, substantial cost savings can ensue in processing, calibration, bandwidth and storage requirements.",
keywords = "calibration, computer vision, image matching, image resolution, image sensors, interpolation, stereo image processing, video signal processing, arrays, cameras, electronics packaging, image processing, PSNR",
author = "{Abd Manap}, Nurulfajar and John Soraghan and Lykourgos Petropoulakis",
year = "2013",
month = "10",
day = "10",
doi = "10.1109/ICSIPA.2013.6707980",
language = "English",
isbn = "9781479902675",
pages = "71--76",
booktitle = "2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA)",
publisher = "IEEE",

}

Abd Manap, N, Soraghan, J & Petropoulakis, L 2013, Multi-level view synthesis (MLVS) based on depth image layer separation (DILS) algorithm for multi-camera view. in 2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA). IEEE, pp. 71-76, 2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA), Meloka, Malaysia, 8/10/13. https://doi.org/10.1109/ICSIPA.2013.6707980

Multi-level view synthesis (MLVS) based on depth image layer separation (DILS) algorithm for multi-camera view. / Abd Manap, Nurulfajar; Soraghan, John; Petropoulakis, Lykourgos.

2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA). IEEE, 2013. p. 71-76.

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

TY - GEN

T1 - Multi-level view synthesis (MLVS) based on depth image layer separation (DILS) algorithm for multi-camera view

AU - Abd Manap, Nurulfajar

AU - Soraghan, John

AU - Petropoulakis, Lykourgos

PY - 2013/10/10

Y1 - 2013/10/10

N2 - A novel Multi-Level View Synthesis (MLVS) approach for 3D vision and free-viewpoint video applications, such as light field imaging, is presented. MLVS exploits the advantages of Depth Image Layer Separation (DILS), a new inter-view interpolation algorithm, by extending stereo to multiple camera configurations. The technique finds the pixel correspondences and synthesis through two levels of matching and synthesis process. The main aim of MLVS is to create a multi-camera view system through a reduced number of actual image acquisition cameras, whilst maintaining the quality of the virtual view synthesis images. The proposed technique is shown to offer improved performance and provide additional views with fewer cameras compared to conventional high volume camera configurations for free-viewpoint video acquisition. Thus, substantial cost savings can ensue in processing, calibration, bandwidth and storage requirements.

AB - A novel Multi-Level View Synthesis (MLVS) approach for 3D vision and free-viewpoint video applications, such as light field imaging, is presented. MLVS exploits the advantages of Depth Image Layer Separation (DILS), a new inter-view interpolation algorithm, by extending stereo to multiple camera configurations. The technique finds the pixel correspondences and synthesis through two levels of matching and synthesis process. The main aim of MLVS is to create a multi-camera view system through a reduced number of actual image acquisition cameras, whilst maintaining the quality of the virtual view synthesis images. The proposed technique is shown to offer improved performance and provide additional views with fewer cameras compared to conventional high volume camera configurations for free-viewpoint video acquisition. Thus, substantial cost savings can ensue in processing, calibration, bandwidth and storage requirements.

KW - calibration

KW - computer vision

KW - image matching

KW - image resolution

KW - image sensors

KW - interpolation

KW - stereo image processing

KW - video signal processing

KW - arrays

KW - cameras

KW - electronics packaging

KW - image processing

KW - PSNR

UR - http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6707980

U2 - 10.1109/ICSIPA.2013.6707980

DO - 10.1109/ICSIPA.2013.6707980

M3 - Conference contribution book

SN - 9781479902675

SP - 71

EP - 76

BT - 2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA)

PB - IEEE

ER -

Abd Manap N, Soraghan J, Petropoulakis L. Multi-level view synthesis (MLVS) based on depth image layer separation (DILS) algorithm for multi-camera view. In 2013 IEEE International Conference on Signal and Image Processing Applications (ICSIPA). IEEE. 2013. p. 71-76 https://doi.org/10.1109/ICSIPA.2013.6707980