电子科技2025，Vol.38Issue(8)：57-65,9.DOI:10.16180/j.cnki.issn1007-7820.2025.08.008
结合并联Transformer和残差U-Net网络的水下图像增强
Underwater Image Enhancement Combining Parallel Transformer and Residual U-Net Networks
摘要
Abstract
In view of the problems of color distortion,low contrast and blurred details in underwater images,an U-Net network based on parallel Transformer and residual convolution is designed for underwater image enhance-ment.In the new U-Net structure,the HCTB(Hybrid Convolution Transformer Block)is placed in the encoding and decoding parts,which integrates the ability of the Transformer to capture global information and the ability of the con-volutional block to capture local information,and builds a number of PAM(Parallel Attention Module)in the hopping connection part to extract more important pixel and channel information.The existing UIEB(Underwater Im-age Enhancement Benchmark dataset)paired dataset is used to train the network.In order to verify the effectiveness of the proposed algorithm,underwater images with different color degree are selected for experiments and tests.The experimental results show that the PSNR(Peak Single-to-Ratio)value of the proposed model is increased by 4.3%compared with other advanced models,and the subjective and objective evaluation results are obtained,which effec-tively improves the enhancement level of underwater images.关键词
水下图像增强/Transformer/残差卷积/U-Net网络/平行注意模块/通道注意/像素注意/卷积神经网络/深度学习Key words
underwater image enhancement/Transformer/residual convolution/U-Net network/parallel attention module/channel attention/pixel attention/convolutional neural networks/deep learning分类
信息技术与安全科学引用本文复制引用
陈清江,李宗莹..结合并联Transformer和残差U-Net网络的水下图像增强[J].电子科技,2025,38(8):57-65,9.基金项目
国家自然科学基金(12202332) （12202332）
陕西省自然科学基础研究计划(2021JQ-495) National Natural Science Foundation of China(12202332) （2021JQ-495）
Shaanxi Provincial Natural Science Basic Research Program(2021JQ-495) （2021JQ-495）
