Distributed Compressed Video Sensing with Joint Optimization of Dictionary Learning and <I>l</I><SUB>1</SUB>-Analysis Based Reconstruction

Fang TIAN; Jie GUO; Bin SONG; Haixiao LIU; Hao QIN

doi:10.1587/transinf.2015EDP7373

Distributed Compressed Video Sensing with Joint Optimization of Dictionary Learning and l₁-Analysis Based Reconstruction

Fang TIAN, Jie GUO, Bin SONG, Haixiao LIU, Hao QIN

Full Text Views

0

Share
Cite this

Summary :

Distributed compressed video sensing (DCVS), combining advantages of compressed sensing and distributed video coding, is developed as a novel and powerful system to get an encoder with low complexity. Nevertheless, it is still unclear how to explore the method to achieve an effective video recovery through utilizing realistic signal characteristics as much as possible. Based on this, we present a novel spatiotemporal dictionary learning (DL) based reconstruction method for DCVS, where both the DL model and the l₁-analysis based recovery with correlation constraints are included in the minimization problem to achieve the joint optimization of sparse representation and signal reconstruction. Besides, an alternating direction method with multipliers (ADMM) based numerical algorithm is outlined for solving the underlying optimization problem. Simulation results demonstrate that the proposed method outperforms other methods, with 0.03-4.14 dB increases in PSNR and a 0.13-15.31 dB gain for non-key frames.

Publication: IEICE TRANSACTIONS on Information Vol.E99-D No.4 pp.1202-1211

Publication Date: 2016/04/01

Publicized: 2016/01/21

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2015EDP7373

Type of Manuscript: PAPER

Category: Image Processing and Video Processing

Authors

Fang TIAN
  Xidian University
Jie GUO
  Xidian University
Bin SONG
  Xidian University
Haixiao LIU
  Xidian University
Hao QIN
  Xidian University

Keyword

compressed video sensing, realistic signal characteristics, dictionary learning, alternating direction method with multipliers

Cite this

Copy

Fang TIAN, Jie GUO, Bin SONG, Haixiao LIU, Hao QIN, "Distributed Compressed Video Sensing with Joint Optimization of Dictionary Learning and l1-Analysis Based Reconstruction" in IEICE TRANSACTIONS on Information, vol. E99-D, no. 4, pp. 1202-1211, April 2016, doi: 10.1587/transinf.2015EDP7373.
Abstract: Distributed compressed video sensing (DCVS), combining advantages of compressed sensing and distributed video coding, is developed as a novel and powerful system to get an encoder with low complexity. Nevertheless, it is still unclear how to explore the method to achieve an effective video recovery through utilizing realistic signal characteristics as much as possible. Based on this, we present a novel spatiotemporal dictionary learning (DL) based reconstruction method for DCVS, where both the DL model and the l₁-analysis based recovery with correlation constraints are included in the minimization problem to achieve the joint optimization of sparse representation and signal reconstruction. Besides, an alternating direction method with multipliers (ADMM) based numerical algorithm is outlined for solving the underlying optimization problem. Simulation results demonstrate that the proposed method outperforms other methods, with 0.03-4.14 dB increases in PSNR and a 0.13-15.31 dB gain for non-key frames.
URL: https://globals.ieice.org/en_transactions/information/10.1587/transinf.2015EDP7373/_p

Copy

@ARTICLE{e99-d_4_1202,
author={Fang TIAN, Jie GUO, Bin SONG, Haixiao LIU, Hao QIN, },
journal={IEICE TRANSACTIONS on Information},
title={Distributed Compressed Video Sensing with Joint Optimization of Dictionary Learning and l1-Analysis Based Reconstruction},
year={2016},
volume={E99-D},
number={4},
pages={1202-1211},
abstract={Distributed compressed video sensing (DCVS), combining advantages of compressed sensing and distributed video coding, is developed as a novel and powerful system to get an encoder with low complexity. Nevertheless, it is still unclear how to explore the method to achieve an effective video recovery through utilizing realistic signal characteristics as much as possible. Based on this, we present a novel spatiotemporal dictionary learning (DL) based reconstruction method for DCVS, where both the DL model and the l₁-analysis based recovery with correlation constraints are included in the minimization problem to achieve the joint optimization of sparse representation and signal reconstruction. Besides, an alternating direction method with multipliers (ADMM) based numerical algorithm is outlined for solving the underlying optimization problem. Simulation results demonstrate that the proposed method outperforms other methods, with 0.03-4.14 dB increases in PSNR and a 0.13-15.31 dB gain for non-key frames.},
keywords={},
doi={10.1587/transinf.2015EDP7373},
ISSN={1745-1361},
month={April},}

Copy

TY - JOUR
TI - Distributed Compressed Video Sensing with Joint Optimization of Dictionary Learning and l1-Analysis Based Reconstruction
T2 - IEICE TRANSACTIONS on Information
SP - 1202
EP - 1211
AU - Fang TIAN
AU - Jie GUO
AU - Bin SONG
AU - Haixiao LIU
AU - Hao QIN
PY - 2016
DO - 10.1587/transinf.2015EDP7373
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E99-D
IS - 4
JA - IEICE TRANSACTIONS on Information
Y1 - April 2016
AB - Distributed compressed video sensing (DCVS), combining advantages of compressed sensing and distributed video coding, is developed as a novel and powerful system to get an encoder with low complexity. Nevertheless, it is still unclear how to explore the method to achieve an effective video recovery through utilizing realistic signal characteristics as much as possible. Based on this, we present a novel spatiotemporal dictionary learning (DL) based reconstruction method for DCVS, where both the DL model and the l₁-analysis based recovery with correlation constraints are included in the minimization problem to achieve the joint optimization of sparse representation and signal reconstruction. Besides, an alternating direction method with multipliers (ADMM) based numerical algorithm is outlined for solving the underlying optimization problem. Simulation results demonstrate that the proposed method outperforms other methods, with 0.03-4.14 dB increases in PSNR and a 0.13-15.31 dB gain for non-key frames.
ER -