基于动态亮度与感受野扩展的阴影去除方法
Journal: Advances in Computer and Autonomous Intelligence Research DOI: 10.12238/acair.v3i1.11931
Abstract
为解决图像部分区域由于阴影覆盖而导致图像质量差及信息不足的问题,对基于动态亮度调整与感受野扩展的阴影去除方法进行了研究,提出了端到端深度网络设计策略。首先结合动态范围扩展和亮度调整技术,提升了阴影区域的清晰度和准确性；其次,通过感受野扩展技术,捕捉图像中的全局信息,解决局部信息不足和阴影边界伪影的问题；最后,通过输出投影获得去除阴影后的图像,并在ISTD+和SRD数据集上进行了验证,结果表明其相较于其他阴影去除方法有显著提升,保留了图像的关键细节和自然纹理。
Keywords
阴影去除；图像增强；深度学习；动态范围扩展；感受野扩展
Full Text
PDF - Viewed/Downloaded: 0 TimesReferences
[1] 徐梦溪,施建强,郑胜男,等.模拟复眼视叶神经网的目标运动方向检测模型[J].智能系统学报,2024,19(3):546-555.
[2] 张歆羽,杨钟亮,周哲画.面向多目标医疗垃圾分类的智能识别分拣系统设计[J].智能系统学报,2024,19(3):584-597.
[3] Le H,Samaras D.Shadow removal via shadow image decomp osition[C].IEEE,2019,8578-8587.
[4] Wang P, Zhao X. Multi-scale adaptive light propagation for shadow reduction[C].IEEE/CVF,2020,5032-5039.
[5] Guo L,Huang S,Liu D,et al. ShadowFormer:Global Context Helps Image Shadow Removal[J].AAAI,2023.
[6] 刘波,田广粮,肖斌,等.利用自适应光照初始化的弱光图像增强方法[J].电子与信息学报,2024,46(02):643-651.
[7] Liu Z,Yin H,Wu X,et al. From shadow generation to shadow removal[C].IEEE/CVF,2021,4927-4936.
[8] 童敢,黄立波,吕雅帅.面向现代GPU的Winograd卷积加速研究[J].电子学报,2024,52(01):244-257.
[9] Hu X,Fu C,Zhu L,et al. Direction-Aware Spatial Context Features for Shadow Detection and Removal[J].IEEE,2020,42(11):2795-2808.
[10] Fu L,Zhou C,Guo Q,et al. Auto-exposure fusion for single image shadow removal[C].IEEE,2021,10571-10580.
[11] Jin Y,Sharma A,Tan R T.Dc-shadownet:Single-image hard and soft shadow removal using unsupervised domainclassifier guided network[C].ICCV,2021.
[12] Cun X,Pun C M,Shi C.Towards Ghost-free Shadow Remov al via Dual Hierarchical Aggregation Network and Shadow Matti ng GAN[J].arXiv,2019,1911.08718.
[13] Zhang H, Xu Y, Zhu S, et al. STC-GAN: Shadow transfer and removal via a generative adversarial network[J].IEEE,2022,31:3844-3857.
[14] Zhu Y, Xiao Z, Fang Y, et al. Efficient model-driven network for shadow removal[C].AAAI,2022,36(3):3635-3643.
[15] Le H, Samaras D. From shadow segmentation to shadow removal[C].IEEE,2020.
[16] YU Qianhao,ZHENG Naishan, HUANG Jie, et al. CNSNet: A Cleanness-Navigated-Shadow Network for Shadow Removal[J]. European Conference on Computer Vision. Cham: Springer Natu re Switzerland,2022,221-238.
[17] ZHANG Xiaofeng,ZHAO Yudi,GU Chaochen,et al.SpA-Form er:An Effective and lightweight Transformer for image shadow removal[J].In IJCNN,2023.
[2] 张歆羽,杨钟亮,周哲画.面向多目标医疗垃圾分类的智能识别分拣系统设计[J].智能系统学报,2024,19(3):584-597.
[3] Le H,Samaras D.Shadow removal via shadow image decomp osition[C].IEEE,2019,8578-8587.
[4] Wang P, Zhao X. Multi-scale adaptive light propagation for shadow reduction[C].IEEE/CVF,2020,5032-5039.
[5] Guo L,Huang S,Liu D,et al. ShadowFormer:Global Context Helps Image Shadow Removal[J].AAAI,2023.
[6] 刘波,田广粮,肖斌,等.利用自适应光照初始化的弱光图像增强方法[J].电子与信息学报,2024,46(02):643-651.
[7] Liu Z,Yin H,Wu X,et al. From shadow generation to shadow removal[C].IEEE/CVF,2021,4927-4936.
[8] 童敢,黄立波,吕雅帅.面向现代GPU的Winograd卷积加速研究[J].电子学报,2024,52(01):244-257.
[9] Hu X,Fu C,Zhu L,et al. Direction-Aware Spatial Context Features for Shadow Detection and Removal[J].IEEE,2020,42(11):2795-2808.
[10] Fu L,Zhou C,Guo Q,et al. Auto-exposure fusion for single image shadow removal[C].IEEE,2021,10571-10580.
[11] Jin Y,Sharma A,Tan R T.Dc-shadownet:Single-image hard and soft shadow removal using unsupervised domainclassifier guided network[C].ICCV,2021.
[12] Cun X,Pun C M,Shi C.Towards Ghost-free Shadow Remov al via Dual Hierarchical Aggregation Network and Shadow Matti ng GAN[J].arXiv,2019,1911.08718.
[13] Zhang H, Xu Y, Zhu S, et al. STC-GAN: Shadow transfer and removal via a generative adversarial network[J].IEEE,2022,31:3844-3857.
[14] Zhu Y, Xiao Z, Fang Y, et al. Efficient model-driven network for shadow removal[C].AAAI,2022,36(3):3635-3643.
[15] Le H, Samaras D. From shadow segmentation to shadow removal[C].IEEE,2020.
[16] YU Qianhao,ZHENG Naishan, HUANG Jie, et al. CNSNet: A Cleanness-Navigated-Shadow Network for Shadow Removal[J]. European Conference on Computer Vision. Cham: Springer Natu re Switzerland,2022,221-238.
[17] ZHANG Xiaofeng,ZHAO Yudi,GU Chaochen,et al.SpA-Form er:An Effective and lightweight Transformer for image shadow removal[J].In IJCNN,2023.
Copyright © 2025 吴博文, 王圣亚
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
