Visibility enhancement of sand-dust images based on depthwise separable convolution

Chunmei Chen, Chenyu Xu, Chang Wang, Shien Yang, Weiguo Zhang, Yakui Liu

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

To address the visual interference problem in 'sand-blind' environments, where sand images suffer from color shifts and severe loss of background detail, we propose a sand image enhancement algorithm based on color correction and depthwise separable convolution. First, the color correction module preprocesses the synthetic sand images to correct color deviations. Then, the SMU (Separable Multi-scale Unit) is applied within the decoupling reconstruction module to address gradient vanishing and extract degradation information. Finally, aggregated depthwise separable convolutions are added to the denoising module to learn textures, thereby obtaining more effective information and target features. Experiments on synthetic datasets demonstrate that, compared to mainstream algorithms such as TFIO, GDCP, AWC, and HRDCP, our algorithm achieves at least 10.46% and 35.56% improvements in PSNR and SSIM metrics, respectively, better restoring image details obscured by sand.

Original languageEnglish
Title of host publication2024 3rd International Conference on Robotics, Artificial Intelligence and Intelligent Control, RAIIC 2024
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages423-429
Number of pages7
ISBN (Electronic)9798350352627
DOIs
StatePublished - 2024
Externally publishedYes
Event3rd International Conference on Robotics, Artificial Intelligence and Intelligent Control, RAIIC 2024 - Hybrid, Mianyang, China
Duration: 5 Jul 20247 Jul 2024

Publication series

Name2024 3rd International Conference on Robotics, Artificial Intelligence and Intelligent Control, RAIIC 2024

Conference

Conference3rd International Conference on Robotics, Artificial Intelligence and Intelligent Control, RAIIC 2024
Country/TerritoryChina
CityHybrid, Mianyang
Period5/07/247/07/24

Keywords

  • 'sand-blind' environment
  • Color correction
  • decoupling reconstruction
  • depthwise separable convolution
  • detail information loss

Fingerprint

Dive into the research topics of 'Visibility enhancement of sand-dust images based on depthwise separable convolution'. Together they form a unique fingerprint.

Cite this