Chen, Y, Wen, Z, Li, P, Zhang, G
ORCID: 0000-0002-2351-2661, Guo, J, Dong, M and Ma, H
(2026)
A DARU-Net-based method for digital holographic phase denoising.
Optics Communications, 624.
ISSN 0030-4018
Preview |
Text
A DARU-Net-based method for digital holographic phase denoising.pdf - Accepted Version Available under License Creative Commons Attribution. Download (2MB) | Preview |
Abstract
In digital holographic measurement, noise is inevitably introduced during the acquisition process, which severely degrades phase quality and hinders accurate recovery of the true phase and three-dimensional morphology of the specimen. Thus, a phase denoising method for digital holography based on dilated attention residual U-Net (DARU-Net) is proposed in this paper. This method adopts U-Net as the backbone and incorporates multiple dilated convolutional blocks (DCBs) with varying dilation rates to enlarge the receptive field, thereby enabling multi-scale feature extraction. A hybrid attention block, combining channel and spatial attention, is integrated into the skip connections to highlight key phase information via adaptive learning of feature weights. A simulated dataset containing various fringe densities and noise levels is constructed for model training and evaluation. In addition, a transmission digital holographic optical path imaging system is constructed to collect holograms of the specimens for experimental validation. Both simulation and experimental results demonstrate that the proposed method outperforms other comparison methods in terms of noise reduction performance across all evaluation metrics, and significantly improves phase unwrapping accuracy. Overall, the proposed method effectively suppresses noise interference and exhibits notable advantages in noise robustness and phase fidelity.
| Item Type: | Article |
|---|---|
| Uncontrolled Keywords: | 4006 Communications Engineering; 40 Engineering; Networking and Information Technology R&D (NITRD); 0205 Optical Physics; 0906 Electrical and Electronic Engineering; 1005 Communications Technologies; Optoelectronics & Photonics; 4006 Communications engineering; 5102 Atomic, molecular and optical physics; 5108 Quantum physics |
| Subjects: | T Technology > TK Electrical engineering. Electronics. Nuclear engineering |
| Divisions: | Engineering and Built Environment |
| Publisher: | Elsevier |
| Date of acceptance: | 10 November 2026 |
| Date of first compliant Open Access: | 17 November 2026 |
| Date Deposited: | 16 Nov 2026 09:48 |
| Last Modified: | 17 Nov 2026 00:50 |
| DOI or ID number: | 10.1016/j.optcom.2026.133979 |
| URI: | https://researchonline.ljmu.ac.uk/id/eprint/29875 |
![]() |
View Item |
Export Citation
Export Citation