Super-resolution video enhancement based on a constrained set of motion vectors

Zoran A. Ivanovski, Lina Karam, Glen P. Abousleman

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

1 Citation (Scopus)

Abstract

Modern video surveillance and target tracking applications utilize multiple cameras transmitting low-bit-rate video through channels of very limited bandwidth. The highly compressed video exhibits coding artifacts that can cause target detection and tracking procedures to fail. Thus, to lower the level of noise and retain the sharpness of the video frames, super-resolution techniques can be employed for video enhancement. In this paper, we propose an efficient super-resolution video enhancement scheme that is based on a constrained set of motion vectors. The proposed scheme computes the motion vectors using the original (uncompressed) video frames, and transmits only a small set of these vectors to the receiver. At the receiver, each pixel is assigned a motion vector from the constrained set to maximize the motion prediction performance. The size of the transmitted vector set is constrained to be less than 3% of the total coded bit stream. In the video enhancement process, an L2-norm minimization super-resolution procedure is applied. The proposed scheme is applied to enhance highly compressed, real-world video sequences. The results obtained show significant improvement in the visual quality of the video sequences, as well as in the performance of subsequent target detection and tracking procedures.

Original languageEnglish (US)
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
EditorsZ. Rahman, R.A. Schowengerdt, S.E. Reichenbach
Pages124-132
Number of pages9
Volume5817
DOIs
StatePublished - 2005
EventVisual Information Processing XIV - Orlando, FL, United States
Duration: Mar 29 2005Mar 30 2005

Other

OtherVisual Information Processing XIV
CountryUnited States
CityOrlando, FL
Period3/29/053/30/05

Fingerprint

Target tracking
augmentation
receivers
video compression
performance prediction
sharpness
surveillance
Image coding
norms
artifacts
coding
Pixels
Cameras
pixels
cameras
bandwidth
Bandwidth
optimization
causes

Keywords

  • Compressed video
  • Motion estimation
  • Motion prediction
  • Super resolution
  • Target tracking
  • Video enhancement
  • Video surveillance

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Condensed Matter Physics

Cite this

Ivanovski, Z. A., Karam, L., & Abousleman, G. P. (2005). Super-resolution video enhancement based on a constrained set of motion vectors. In Z. Rahman, R. A. Schowengerdt, & S. E. Reichenbach (Eds.), Proceedings of SPIE - The International Society for Optical Engineering (Vol. 5817, pp. 124-132). [13] https://doi.org/10.1117/12.606625

Super-resolution video enhancement based on a constrained set of motion vectors. / Ivanovski, Zoran A.; Karam, Lina; Abousleman, Glen P.

Proceedings of SPIE - The International Society for Optical Engineering. ed. / Z. Rahman; R.A. Schowengerdt; S.E. Reichenbach. Vol. 5817 2005. p. 124-132 13.

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

Ivanovski, ZA, Karam, L & Abousleman, GP 2005, Super-resolution video enhancement based on a constrained set of motion vectors. in Z Rahman, RA Schowengerdt & SE Reichenbach (eds), Proceedings of SPIE - The International Society for Optical Engineering. vol. 5817, 13, pp. 124-132, Visual Information Processing XIV, Orlando, FL, United States, 3/29/05. https://doi.org/10.1117/12.606625
Ivanovski ZA, Karam L, Abousleman GP. Super-resolution video enhancement based on a constrained set of motion vectors. In Rahman Z, Schowengerdt RA, Reichenbach SE, editors, Proceedings of SPIE - The International Society for Optical Engineering. Vol. 5817. 2005. p. 124-132. 13 https://doi.org/10.1117/12.606625
Ivanovski, Zoran A. ; Karam, Lina ; Abousleman, Glen P. / Super-resolution video enhancement based on a constrained set of motion vectors. Proceedings of SPIE - The International Society for Optical Engineering. editor / Z. Rahman ; R.A. Schowengerdt ; S.E. Reichenbach. Vol. 5817 2005. pp. 124-132
@inproceedings{2fb5f9b0933240fe89542dac6ceb7256,
title = "Super-resolution video enhancement based on a constrained set of motion vectors",
abstract = "Modern video surveillance and target tracking applications utilize multiple cameras transmitting low-bit-rate video through channels of very limited bandwidth. The highly compressed video exhibits coding artifacts that can cause target detection and tracking procedures to fail. Thus, to lower the level of noise and retain the sharpness of the video frames, super-resolution techniques can be employed for video enhancement. In this paper, we propose an efficient super-resolution video enhancement scheme that is based on a constrained set of motion vectors. The proposed scheme computes the motion vectors using the original (uncompressed) video frames, and transmits only a small set of these vectors to the receiver. At the receiver, each pixel is assigned a motion vector from the constrained set to maximize the motion prediction performance. The size of the transmitted vector set is constrained to be less than 3{\%} of the total coded bit stream. In the video enhancement process, an L2-norm minimization super-resolution procedure is applied. The proposed scheme is applied to enhance highly compressed, real-world video sequences. The results obtained show significant improvement in the visual quality of the video sequences, as well as in the performance of subsequent target detection and tracking procedures.",
keywords = "Compressed video, Motion estimation, Motion prediction, Super resolution, Target tracking, Video enhancement, Video surveillance",
author = "Ivanovski, {Zoran A.} and Lina Karam and Abousleman, {Glen P.}",
year = "2005",
doi = "10.1117/12.606625",
language = "English (US)",
volume = "5817",
pages = "124--132",
editor = "Z. Rahman and R.A. Schowengerdt and S.E. Reichenbach",
booktitle = "Proceedings of SPIE - The International Society for Optical Engineering",

}

TY - GEN

T1 - Super-resolution video enhancement based on a constrained set of motion vectors

AU - Ivanovski, Zoran A.

AU - Karam, Lina

AU - Abousleman, Glen P.

PY - 2005

Y1 - 2005

N2 - Modern video surveillance and target tracking applications utilize multiple cameras transmitting low-bit-rate video through channels of very limited bandwidth. The highly compressed video exhibits coding artifacts that can cause target detection and tracking procedures to fail. Thus, to lower the level of noise and retain the sharpness of the video frames, super-resolution techniques can be employed for video enhancement. In this paper, we propose an efficient super-resolution video enhancement scheme that is based on a constrained set of motion vectors. The proposed scheme computes the motion vectors using the original (uncompressed) video frames, and transmits only a small set of these vectors to the receiver. At the receiver, each pixel is assigned a motion vector from the constrained set to maximize the motion prediction performance. The size of the transmitted vector set is constrained to be less than 3% of the total coded bit stream. In the video enhancement process, an L2-norm minimization super-resolution procedure is applied. The proposed scheme is applied to enhance highly compressed, real-world video sequences. The results obtained show significant improvement in the visual quality of the video sequences, as well as in the performance of subsequent target detection and tracking procedures.

AB - Modern video surveillance and target tracking applications utilize multiple cameras transmitting low-bit-rate video through channels of very limited bandwidth. The highly compressed video exhibits coding artifacts that can cause target detection and tracking procedures to fail. Thus, to lower the level of noise and retain the sharpness of the video frames, super-resolution techniques can be employed for video enhancement. In this paper, we propose an efficient super-resolution video enhancement scheme that is based on a constrained set of motion vectors. The proposed scheme computes the motion vectors using the original (uncompressed) video frames, and transmits only a small set of these vectors to the receiver. At the receiver, each pixel is assigned a motion vector from the constrained set to maximize the motion prediction performance. The size of the transmitted vector set is constrained to be less than 3% of the total coded bit stream. In the video enhancement process, an L2-norm minimization super-resolution procedure is applied. The proposed scheme is applied to enhance highly compressed, real-world video sequences. The results obtained show significant improvement in the visual quality of the video sequences, as well as in the performance of subsequent target detection and tracking procedures.

KW - Compressed video

KW - Motion estimation

KW - Motion prediction

KW - Super resolution

KW - Target tracking

KW - Video enhancement

KW - Video surveillance

UR - http://www.scopus.com/inward/record.url?scp=27644554542&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=27644554542&partnerID=8YFLogxK

U2 - 10.1117/12.606625

DO - 10.1117/12.606625

M3 - Conference contribution

VL - 5817

SP - 124

EP - 132

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

A2 - Rahman, Z.

A2 - Schowengerdt, R.A.

A2 - Reichenbach, S.E.

ER -