Imaging through fluctuating scattering media such as fog is of challenge since it seriously degrades the image quality.We investigate how the image quality of computational ghost imaging is reduced by fluctuating fog ...Imaging through fluctuating scattering media such as fog is of challenge since it seriously degrades the image quality.We investigate how the image quality of computational ghost imaging is reduced by fluctuating fog and how to obtain a high-quality defogging ghost image. We show theoretically and experimentally that the photon number fluctuations introduced by fluctuating fog is the reason for ghost image degradation. An algorithm is proposed to process the signals collected by the computational ghost imaging device to eliminate photon number fluctuations of different measurement events. Thus, a high-quality defogging ghost image is reconstructed even though fog is evenly distributed on the optical path. A nearly 100% defogging ghost image is obtained by further using a cycle generative adversarial network to process the reconstructed defogging image.展开更多
心脏磁共振成像(cardiac magnetic resonance,CMR)过程中患者误动、异常幅度的呼吸运动、心律失常会造成CMR图像质量下降,为解决现有的CMR图像增强网络需要人为制作配对数据,且图像增强后部分组织纹理细节丢失的问题,提出了基于空频域...心脏磁共振成像(cardiac magnetic resonance,CMR)过程中患者误动、异常幅度的呼吸运动、心律失常会造成CMR图像质量下降,为解决现有的CMR图像增强网络需要人为制作配对数据,且图像增强后部分组织纹理细节丢失的问题,提出了基于空频域特征学习的循环一致性生成对抗网络(cycle-consistent generative adversavial network based on spatial-frequency domain feature learning,SFFL-CycleGAN).研究结果表明,该网络无须人为制作配对数据集,增强后的CMR图像组织纹理细节丰富,在结构相似度(structural similarity,SSIM)和峰值信噪比(peak signal to noise ratio,PSNR)等方面均优于现有的配对训练网络以及原始的CycleGAN网络,图像增强效果好,有效助力病情诊断.展开更多
Side-scan sonar(SSS)is essential for acquiring high-resolution seafloor images over large areas,facilitat-ing the identification of subsea objects.However,military security restrictions and the scarcity of subsea targ...Side-scan sonar(SSS)is essential for acquiring high-resolution seafloor images over large areas,facilitat-ing the identification of subsea objects.However,military security restrictions and the scarcity of subsea targets limit the availability of SSS data,posing challenges for Automatic Target Recognition(ATR)research.This paper presents an approach that uses Cycle-Consistent Generative Adversarial Networks(CycleGAN)to augment SSS images of key subsea objects,such as shipwrecks,aircraft,and drowning victims.The process begins by constructing 3D models to generate rendered images with realistic shadows frommultiple angles.To enhance image quality,a shadowextractor and shadow region loss function are introduced to ensure consistent shadowrepresentation.Additionally,amulti-resolution learning structure enables effective training,even with limited data availability.The experimental results show that the generated data improved object detection accuracy when they were used for training and demonstrated the ability to generate clear shadow and background regions with stability.展开更多
建筑平面图外墙的自动识别在建筑设计和施工的过程中具有重要意义,它不仅能提高设计效率,还能辅助设计审核,降低图纸错误风险。同时,自动识别外墙等构件为三维模型的快速生成提供有效数据,推动建筑信息建模(Building Information Modeli...建筑平面图外墙的自动识别在建筑设计和施工的过程中具有重要意义,它不仅能提高设计效率,还能辅助设计审核,降低图纸错误风险。同时,自动识别外墙等构件为三维模型的快速生成提供有效数据,推动建筑信息建模(Building Information Modeling,BIM)技术的发展。目前自动识别建筑平面图外墙的方法主要是基于规则的方法,且其效果仍有一定的提升空间。针对上述问题,本文提出了一种基于Swin Transformer和LSTM的CycleGAN算法实现建筑平面图外墙的自动识别,以辅助建筑平面图的设计、建筑平面图设计的质量稽查、三维模型的快速生成等。该算法基于循环一致性生成对抗网络(CycleGAN)框架,嵌入Swin Transformer Block模块和长短期记忆(Long Short-Term Memory,LSTM)模块,增强模型特征提取能力和网络学习能力。实验结果表明,改进的CycleGAN算法实现了非成对训练样本条件下建筑平面图外墙的识别,在结构相似性指数和峰值信噪比的两个指标上分别提高了22.4%和4.223,且生成的建筑外墙图具有清晰度高、细节特征完整的特点。展开更多
基金supported by the Natural Science Foundation of Shandong Province, China (Grant No. ZR2022MF249)。
文摘Imaging through fluctuating scattering media such as fog is of challenge since it seriously degrades the image quality.We investigate how the image quality of computational ghost imaging is reduced by fluctuating fog and how to obtain a high-quality defogging ghost image. We show theoretically and experimentally that the photon number fluctuations introduced by fluctuating fog is the reason for ghost image degradation. An algorithm is proposed to process the signals collected by the computational ghost imaging device to eliminate photon number fluctuations of different measurement events. Thus, a high-quality defogging ghost image is reconstructed even though fog is evenly distributed on the optical path. A nearly 100% defogging ghost image is obtained by further using a cycle generative adversarial network to process the reconstructed defogging image.
文摘心脏磁共振成像(cardiac magnetic resonance,CMR)过程中患者误动、异常幅度的呼吸运动、心律失常会造成CMR图像质量下降,为解决现有的CMR图像增强网络需要人为制作配对数据,且图像增强后部分组织纹理细节丢失的问题,提出了基于空频域特征学习的循环一致性生成对抗网络(cycle-consistent generative adversavial network based on spatial-frequency domain feature learning,SFFL-CycleGAN).研究结果表明,该网络无须人为制作配对数据集,增强后的CMR图像组织纹理细节丰富,在结构相似度(structural similarity,SSIM)和峰值信噪比(peak signal to noise ratio,PSNR)等方面均优于现有的配对训练网络以及原始的CycleGAN网络,图像增强效果好,有效助力病情诊断.
基金supported by the National Research Foundation of Korea(NRF)grant funded by the Korea government(MSIT)(No.RS-2024-00334159)the Korea Institute of Ocean Science and Technology(KIOST)project entitled“Development of Maritime Domain Awareness Technology for Sea Power Enhancement”(PEA0332).
文摘Side-scan sonar(SSS)is essential for acquiring high-resolution seafloor images over large areas,facilitat-ing the identification of subsea objects.However,military security restrictions and the scarcity of subsea targets limit the availability of SSS data,posing challenges for Automatic Target Recognition(ATR)research.This paper presents an approach that uses Cycle-Consistent Generative Adversarial Networks(CycleGAN)to augment SSS images of key subsea objects,such as shipwrecks,aircraft,and drowning victims.The process begins by constructing 3D models to generate rendered images with realistic shadows frommultiple angles.To enhance image quality,a shadowextractor and shadow region loss function are introduced to ensure consistent shadowrepresentation.Additionally,amulti-resolution learning structure enables effective training,even with limited data availability.The experimental results show that the generated data improved object detection accuracy when they were used for training and demonstrated the ability to generate clear shadow and background regions with stability.
