RepDNet:A re-parameterization despeckling network for autonomous underwater side-scan sonar imaging with prior-knowledge customized convolution  

作　　者：Zhuoyi Li Zhisen Wang Deshan Chen Tsz Leung Yip Angelo P.Teixeira 

机构地区：[1]Department of Logistics and Maritime Studies,The Hong Kong Polytechnic University,China [2]State Key Laboratory of Maritime Technology and Safety,Wuhan University of Technology,China [3]School of Transportation and Logistics Engineering,Wuhan University of Technology,China [4]National Engineering Research Center for Water Transport Safety,Wuhan University of Technology,China [5]Centre for Marine Technology and Ocean Engineering(CENTEC),Instituto Superior Tecnico,Universidade de Lisboa,Portugal

出　　处：《Defence Technology（防务技术）》2024年第5期259-274,共16页Defence Technology

基　　金：supported by the National Key R&D Program of China(Grant No.2023YFC3010803);the National Nature Science Foundation of China(Grant No.52272424);the Key R&D Program of Hubei Province of China(Grant No.2023BCB123);the Fundamental Research Funds for the Central Universities(Grant No.WUT:2023IVB079)。

摘　　要：Side-scan sonar(SSS)is now a prevalent instrument for large-scale seafloor topography measurements,deployable on an autonomous underwater vehicle(AUV)to execute fully automated underwater acoustic scanning imaging along a predetermined trajectory.However,SSS images often suffer from speckle noise caused by mutual interference between echoes,and limited AUV computational resources further hinder noise suppression.Existing approaches for SSS image processing and speckle noise reduction rely heavily on complex network structures and fail to combine the benefits of deep learning and domain knowledge.To address the problem,Rep DNet,a novel and effective despeckling convolutional neural network is proposed.Rep DNet introduces two re-parameterized blocks:the Pixel Smoothing Block(PSB)and Edge Enhancement Block(EEB),preserving edge information while attenuating speckle noise.During training,PSB and EEB manifest as double-layered multi-branch structures,integrating first-order and secondorder derivatives and smoothing functions.During inference,the branches are re-parameterized into a 3×3 convolution,enabling efficient inference without sacrificing accuracy.Rep DNet comprises three computational operations:3×3 convolution,element-wise summation and Rectified Linear Unit activation.Evaluations on benchmark datasets,a real SSS dataset and Data collected at Lake Mulan aestablish Rep DNet as a well-balanced network,meeting the AUV computational constraints in terms of performance and latency.

关 键 词：Side-scan sonar Sonar image despeckling Domain knowledge RE-PARAMETERIZATION 

分 类 号：U666.7[交通运输工程—船舶及航道工程]