Low-light image enhancement aims to improve the visibility of severely degraded images captured under insufficient illumination,alleviating the adverse effects of illumination degradation on image quality.Traditional ...Low-light image enhancement aims to improve the visibility of severely degraded images captured under insufficient illumination,alleviating the adverse effects of illumination degradation on image quality.Traditional Retinex-based approaches,inspired by human visual perception of brightness and color,decompose an image into illumination and reflectance components to restore fine details.However,their limited capacity for handling noise and complex lighting conditions often leads to distortions and artifacts in the enhanced results,particularly under extreme low-light scenarios.Although deep learning methods built upon Retinex theory have recently advanced the field,most still suffer frominsufficient interpretability and sub-optimal enhancement performance.This paper presents RetinexWT,a novel framework that tightly integrates classical Retinex theory with modern deep learning.Following Retinex principles,RetinexWT employs wavelet transforms to estimate illumination maps for brightness adjustment.A detail-recovery module that synergistically combines Vision Transformer(ViT)and wavelet transforms is then introduced to guide the restoration of lost details,thereby improving overall image quality.Within the framework,wavelet decomposition splits input features into high-frequency and low-frequency components,enabling scale-specific processing of global illumination/color cues and fine textures.Furthermore,a gating mechanism selectively fuses down-sampled and up-sampled features,while an attention-based fusion strategy enhances model interpretability.Extensive experiments on the LOL dataset demonstrate that RetinexWT surpasses existing Retinex-oriented deeplearning methods,achieving an average Peak Signal-to-Noise Ratio(PSNR)improvement of 0.22 dB over the current StateOfTheArt(SOTA),thereby confirming its superiority in low-light image enhancement.Code is available at https://github.com/CHEN-hJ516/RetinexWT(accessed on 14 October 2025).展开更多
Photoacoustic-computed tomography is a novel imaging technique that combines high absorption contrast and deep tissue penetration capability,enabling comprehensive three-dimensional imaging of biological targets.Howev...Photoacoustic-computed tomography is a novel imaging technique that combines high absorption contrast and deep tissue penetration capability,enabling comprehensive three-dimensional imaging of biological targets.However,the increasing demand for higher resolution and real-time imaging results in significant data volume,limiting data storage,transmission and processing efficiency of system.Therefore,there is an urgent need for an effective method to compress the raw data without compromising image quality.This paper presents a photoacoustic-computed tomography 3D data compression method and system based on Wavelet-Transformer.This method is based on the cooperative compression framework that integrates wavelet hard coding with deep learning-based soft decoding.It combines the multiscale analysis capability of wavelet transforms with the global feature modeling advantage of Transformers,achieving high-quality data compression and reconstruction.Experimental results using k-wave simulation suggest that the proposed compression system has advantages under extreme compression conditions,achieving a raw data compression ratio of up to 1:40.Furthermore,three-dimensional data compression experiment using in vivo mouse demonstrated that the maximum peak signal-to-noise ratio(PSNR)and structural similarity index(SSIM)values of reconstructed images reached 38.60 and 0.9583,effectively overcoming detail loss and artifacts introduced by raw data compression.All the results suggest that the proposed system can significantly reduce storage requirements and hardware cost,enhancing computational efficiency and image quality.These advantages support the development of photoacoustic-computed tomography toward higher efficiency,real-time performance and intelligent functionality.展开更多
Inspections of power transmission lines(PTLs)conducted using unmanned aerial vehicles(UAVs)are complicated by the fine structure of the lines and complex backgrounds,making accurate and efficient segmentation challeng...Inspections of power transmission lines(PTLs)conducted using unmanned aerial vehicles(UAVs)are complicated by the fine structure of the lines and complex backgrounds,making accurate and efficient segmentation challenging.This study presents the Wavelet-Guided Transformer U-Net(WGT-UNet)model,a new hybrid net-work that combines Convolutional Neural Networks(CNNs),Discrete Wavelet Transform(DWT),and Transformer architectures.The model’s primary contribution is based on spatial and channel attention mechanisms derived from wavelet subbands to guide the Transformer’s self-attention structure.Thus,low and high frequency components are separated at each stage using DWT,suppressing structural noise and making linear objects more prominent.The developed design is supported by multi-component hybrid cost functions that simultaneously solve class imbalance,edge sharpness,structural integrity,and spatial regularity issues.Furthermore,high segmentation success has been achieved in producing sharp boundaries and continuous line structures with the DWT-guided attention mechanism.Experiments conducted on the TTPLA dataset reveal that the version using the ConvNeXt backbone outperforms the current state-of-the-art approaches with an F1-Score of 79.33%and an Intersection over Union(IoU)value of 68.38%.The models and visual outputs of the developed method and all compared models can be accessed at https://github.com/burhanbarakli/WGT-UNET.展开更多
针对地图综合中建筑多边形化简方法依赖人工规则、自动化程度低且难以利用已有化简成果的问题,本文提出了一种基于Transformer机制的建筑多边形化简模型。该模型首先把建筑多边形映射至一定范围的网格空间,将建筑多边形的坐标串表达为...针对地图综合中建筑多边形化简方法依赖人工规则、自动化程度低且难以利用已有化简成果的问题,本文提出了一种基于Transformer机制的建筑多边形化简模型。该模型首先把建筑多边形映射至一定范围的网格空间,将建筑多边形的坐标串表达为网格序列,从而获取建筑多边形化简前后的Token序列,构建出建筑多边形化简样本对数据;随后采用Transformer架构建立模型,基于样本数据利用模型的掩码自注意力机制学习点序列之间的依赖关系,最终逐点生成新的简化多边形,从而实现建筑多边形的化简。在训练过程中,模型使用结构化的样本数据,设计了忽略特定索引的交叉熵损失函数以提升化简质量。试验设计包括主试验与泛化验证两部分。主试验基于洛杉矶1∶2000建筑数据集,分别采用0.2、0.3和0.5 mm 3种网格尺寸对多边形进行编码,实现了目标比例尺为1∶5000与1∶10000的化简。试验结果表明,在0.3 mm的网格尺寸下模型性能最优,验证集上的化简结果与人工标注的一致率超过92.0%,且针对北京部分区域的建筑多边形数据的泛化试验验证了模型的迁移能力;与LSTM模型的对比分析显示,在参数规模相近的条件下,LSTM模型无法形成有效收敛,并生成可用结果。本文证实了Transformer在处理空间几何序列任务中的潜力,且能够有效复用已有化简样本,为智能建筑多边形化简提供了具有工程实用价值的途径。展开更多
基金supported in part by the National Natural Science Foundation of China[Grant number 62471075]the Major Science and Technology Project Grant of the Chongqing Municipal Education Commission[Grant number KJZD-M202301901].
文摘Low-light image enhancement aims to improve the visibility of severely degraded images captured under insufficient illumination,alleviating the adverse effects of illumination degradation on image quality.Traditional Retinex-based approaches,inspired by human visual perception of brightness and color,decompose an image into illumination and reflectance components to restore fine details.However,their limited capacity for handling noise and complex lighting conditions often leads to distortions and artifacts in the enhanced results,particularly under extreme low-light scenarios.Although deep learning methods built upon Retinex theory have recently advanced the field,most still suffer frominsufficient interpretability and sub-optimal enhancement performance.This paper presents RetinexWT,a novel framework that tightly integrates classical Retinex theory with modern deep learning.Following Retinex principles,RetinexWT employs wavelet transforms to estimate illumination maps for brightness adjustment.A detail-recovery module that synergistically combines Vision Transformer(ViT)and wavelet transforms is then introduced to guide the restoration of lost details,thereby improving overall image quality.Within the framework,wavelet decomposition splits input features into high-frequency and low-frequency components,enabling scale-specific processing of global illumination/color cues and fine textures.Furthermore,a gating mechanism selectively fuses down-sampled and up-sampled features,while an attention-based fusion strategy enhances model interpretability.Extensive experiments on the LOL dataset demonstrate that RetinexWT surpasses existing Retinex-oriented deeplearning methods,achieving an average Peak Signal-to-Noise Ratio(PSNR)improvement of 0.22 dB over the current StateOfTheArt(SOTA),thereby confirming its superiority in low-light image enhancement.Code is available at https://github.com/CHEN-hJ516/RetinexWT(accessed on 14 October 2025).
基金supported by the National Key R&D Program of China[Grant No.2023YFF0713600]the National Natural Science Foundation of China[Grant No.62275062]+3 种基金Project of Shandong Innovation and Startup Community of High-end Medical Apparatus and Instruments[Grant No.2023-SGTTXM-002 and 2024-SGTTXM-005]the Shandong Province Technology Innovation Guidance Plan(Central Leading Local Science and Technology Development Fund)[Grant No.YDZX2023115]the Taishan Scholar Special Funding Project of Shandong Provincethe Shandong Laboratory of Advanced Biomaterials and Medical Devices in Weihai[Grant No.ZL202402].
文摘Photoacoustic-computed tomography is a novel imaging technique that combines high absorption contrast and deep tissue penetration capability,enabling comprehensive three-dimensional imaging of biological targets.However,the increasing demand for higher resolution and real-time imaging results in significant data volume,limiting data storage,transmission and processing efficiency of system.Therefore,there is an urgent need for an effective method to compress the raw data without compromising image quality.This paper presents a photoacoustic-computed tomography 3D data compression method and system based on Wavelet-Transformer.This method is based on the cooperative compression framework that integrates wavelet hard coding with deep learning-based soft decoding.It combines the multiscale analysis capability of wavelet transforms with the global feature modeling advantage of Transformers,achieving high-quality data compression and reconstruction.Experimental results using k-wave simulation suggest that the proposed compression system has advantages under extreme compression conditions,achieving a raw data compression ratio of up to 1:40.Furthermore,three-dimensional data compression experiment using in vivo mouse demonstrated that the maximum peak signal-to-noise ratio(PSNR)and structural similarity index(SSIM)values of reconstructed images reached 38.60 and 0.9583,effectively overcoming detail loss and artifacts introduced by raw data compression.All the results suggest that the proposed system can significantly reduce storage requirements and hardware cost,enhancing computational efficiency and image quality.These advantages support the development of photoacoustic-computed tomography toward higher efficiency,real-time performance and intelligent functionality.
文摘Inspections of power transmission lines(PTLs)conducted using unmanned aerial vehicles(UAVs)are complicated by the fine structure of the lines and complex backgrounds,making accurate and efficient segmentation challenging.This study presents the Wavelet-Guided Transformer U-Net(WGT-UNet)model,a new hybrid net-work that combines Convolutional Neural Networks(CNNs),Discrete Wavelet Transform(DWT),and Transformer architectures.The model’s primary contribution is based on spatial and channel attention mechanisms derived from wavelet subbands to guide the Transformer’s self-attention structure.Thus,low and high frequency components are separated at each stage using DWT,suppressing structural noise and making linear objects more prominent.The developed design is supported by multi-component hybrid cost functions that simultaneously solve class imbalance,edge sharpness,structural integrity,and spatial regularity issues.Furthermore,high segmentation success has been achieved in producing sharp boundaries and continuous line structures with the DWT-guided attention mechanism.Experiments conducted on the TTPLA dataset reveal that the version using the ConvNeXt backbone outperforms the current state-of-the-art approaches with an F1-Score of 79.33%and an Intersection over Union(IoU)value of 68.38%.The models and visual outputs of the developed method and all compared models can be accessed at https://github.com/burhanbarakli/WGT-UNET.
文摘针对地图综合中建筑多边形化简方法依赖人工规则、自动化程度低且难以利用已有化简成果的问题,本文提出了一种基于Transformer机制的建筑多边形化简模型。该模型首先把建筑多边形映射至一定范围的网格空间,将建筑多边形的坐标串表达为网格序列,从而获取建筑多边形化简前后的Token序列,构建出建筑多边形化简样本对数据;随后采用Transformer架构建立模型,基于样本数据利用模型的掩码自注意力机制学习点序列之间的依赖关系,最终逐点生成新的简化多边形,从而实现建筑多边形的化简。在训练过程中,模型使用结构化的样本数据,设计了忽略特定索引的交叉熵损失函数以提升化简质量。试验设计包括主试验与泛化验证两部分。主试验基于洛杉矶1∶2000建筑数据集,分别采用0.2、0.3和0.5 mm 3种网格尺寸对多边形进行编码,实现了目标比例尺为1∶5000与1∶10000的化简。试验结果表明,在0.3 mm的网格尺寸下模型性能最优,验证集上的化简结果与人工标注的一致率超过92.0%,且针对北京部分区域的建筑多边形数据的泛化试验验证了模型的迁移能力;与LSTM模型的对比分析显示,在参数规模相近的条件下,LSTM模型无法形成有效收敛,并生成可用结果。本文证实了Transformer在处理空间几何序列任务中的潜力,且能够有效复用已有化简样本,为智能建筑多边形化简提供了具有工程实用价值的途径。
