GBRAMP: A generalized backtracking regularized adaptive matching pursuit algorithm for signal reconstruction

Asogbon, M. G., Lu, Y., Samuel, O. W., Jing, L., Miller, A. A. , Li, G. and Wong, K. K.L. (2021) GBRAMP: A generalized backtracking regularized adaptive matching pursuit algorithm for signal reconstruction. Computers and Electrical Engineering, 92, 107189. (doi: 10.1016/j.compeleceng.2021.107189)

[img] Text
241302.pdf - Accepted Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

1MB

Abstract

In order to resolve the problem of excessive processing time and inadequate accuracy caused by existing algorithms in robot vision image reconstruction, a block variable step size adaptive compression sensor reconstruction algorithm is proposed. The algorithm integrates the regularized orthogonal matching pursuit technique in a seamlessly efficient manner to obtain consistent and accurate signal reconstruction outcomes. To apply this technique, a set of selected atoms is initialized by setting fuzzy threshold. Subsequently, inappropriate atoms are excluded, and an iterative procedure is initiated to update the set so as to approximate the signal sparsity in a stepwise fashion. In comparison with commonly used algorithms, the proposed algorithm achieved the lowest signal recovery and reconstruction error. Findings from this study indicate that our proposed hybrid paradigm may lead to positive advancement towards the development of intelligent robotic vision systems for industrial applications.

Item Type:Articles
Additional Information:The Research work was supported in part by the Natural Science Foundation of Top Talent of Shenzhen Technology University (Grant No. 2019207) and the National Natural Science Foundation of China (Grants No. 82050410452, U1613222, 8201101443). Mojisola Grace Asogbon appreciates the support of Chinese Government Scholarship in the pursuit of a PhD degree at the University of Chinese Academy of Sciences, Beijing, China.
Status:Published
Refereed:Yes
Glasgow Author(s) Enlighten ID:Miller, Professor Alice
Authors: Asogbon, M. G., Lu, Y., Samuel, O. W., Jing, L., Miller, A. A., Li, G., and Wong, K. K.L.
College/School:College of Science and Engineering > School of Computing Science
Journal Name:Computers and Electrical Engineering
Publisher:Elsevier
ISSN:0045-7906
ISSN (Online):1879-0755
Published Online:08 May 2021
Copyright Holders:Copyright © 2021 Elsevier Ltd.
First Published:First published in Computers and Electrical Engineering 92:107189
Publisher Policy:Reproduced in accordance with the publisher copyright policy

University Staff: Request a correction | Enlighten Editors: Update this record