+高级检索
首页 > 过刊浏览>2023年第46卷第18期 >171-178
PDF HTML阅读 XML下载 导出引用 引用提醒
基于改进循环生成对抗网络实现红外图像生成
CSTR:
[cstr]
作者:
作者单位:

1.沈阳化工大学信息工程学院 沈阳 110142; 2.中国科学院光电信息处理重点实验室 沈阳 110016; 3.中国科学院沈阳自动化研究所 沈阳 110016; 4.中国科学院机器人与智能制造创新研究院 沈阳 110169

中图分类号:

TP391

基金项目:

中国国家装备发展部重点预研基金（41401040105）项目资助


Based on improved cycle generative adversarial network for infrared image generation
Author:
  • Yi Xing

    Yi Xing

    1.School of Information Engineering,Shenyang University of Chemical Technology，Shenyang 110142, China； 2.Shenyang Institute of Automation, Chinese Academy of Sciences，Shenyang 110016, China； 3.Key Laboratory of OpticalElectronics Information Technology Processing, Shenyang Institute of Automation, Chinese Academy of Sciences，Shenyang 110016， China； 4.Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences，Shenyang 110169, China
    在期刊界中查找
    在百度中查找
    在本站中查找
  • Pan Hao 1

    Pan Hao

    School of Information Engineering,Shenyang University of Chemical Technology，Shenyang 110142, China
    在期刊界中查找
    在百度中查找
    在本站中查找
  • Zhao Huaici

    Zhao Huaici

    2.Shenyang Institute of Automation, Chinese Academy of Sciences，Shenyang 110016, China； 3.Key Laboratory of OpticalElectronics Information Technology Processing, Shenyang Institute of Automation, Chinese Academy of Sciences，Shenyang 110016， China； 4.Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences，Shenyang 110169, China
    在期刊界中查找
    在百度中查找
    在本站中查找
  • Liu Pengfei

    Liu Pengfei

    2.Shenyang Institute of Automation, Chinese Academy of Sciences，Shenyang 110016, China； 3.Key Laboratory of OpticalElectronics Information Technology Processing, Shenyang Institute of Automation, Chinese Academy of Sciences，Shenyang 110016， China； 4.Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences，Shenyang 110169, China
    在期刊界中查找
    在百度中查找
    在本站中查找
  • Yang Bin 1

    Yang Bin

    School of Information Engineering,Shenyang University of Chemical Technology，Shenyang 110142, China
    在期刊界中查找
    在百度中查找
    在本站中查找
Affiliation:

1.School of Information Engineering,Shenyang University of Chemical Technology，Shenyang 110142, China; 2.Shenyang Institute of Automation, Chinese Academy of Sciences，Shenyang 110016, China; 3.Key Laboratory of OpticalElectronics Information Technology Processing, Shenyang Institute of Automation, Chinese Academy of Sciences，Shenyang 110016， China; 4.Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences，Shenyang 110169, China

  • 摘要
    • | |
  • 访问统计
    • |
  • 参考文献
    • |
  • 相似文献
    • |
  • 引证文献
    • | |
  • 文章评论
    摘要:

    针对目前已有的可见光图像生成红外图像的算法不能感知图像的弱纹理区域而导致生成的图像细节信息不突出、图像质量低的问题，本文提出了一种适用于图像生成任务的改进循环生成对抗网络（CycleGAN）结构。首先，利用特征提取能力更强的残差网络构建CycleGAN的生成器网络结构，使图像特征可以充分被提取，解决图像因特征提取不充分导致图像质量低下的问题；其次，在生成器的网络结构中引入了通道注意力机制和空间注意力机制，利用注意力机制对图像感知能力较差的区域进行权重处理，解决图像纹理细节丢失的问题。在OSU数据集上，本文所提出的方法相较于CycleGAN方法在峰值信噪比(PSNR)以及结构相似性(SSIM)指标上分别提高了7.1%和10.9%，在Flir数据集上的PSNR和SSIM分别提高了4.0%和6.7%。经过多个数据集上的实验结果证明，本文改进的方法能够突出图像生成任务中的细节特征信息，并且能有效地提升图像生成的质量。

    Abstract:

    To address the problem that the existing algorithms for generating infrared images from visible images cannot perceive the weak texture regions of the images, which leads to the low quality of the generated image details, this paper proposes an improved Cycle Generation Adversarial Network (CycleGAN) for the image generation task. structure for image generation tasks. Firstly, the generator network structure of the cycle generation adversarial network is constructed by using the residual network with stronger feature extraction ability, so that the image features can be fully extracted and the problem of low image quality caused by insufficient feature extraction can be solved; secondly, the channel attention mechanism and spatial attention mechanism are introduced in the generator network structure, and the regions with poor image perception are weighted by the attention mechanism to solve the problem of loss of image texture details. processing to solve the problem of image texture detail loss. On the OSU dataset, the proposed method improves the Peak Signal to Noise Ratio (PSNR) and Structural Similarity (SSIM) metrics by 7.1% and 10.9%, respectively, compared with the cyclic generative adversarial network method on the Flir dataset. PSNR and SSIM improved by 4.0% and 6.7%, respectively, on the Flir dataset. The experimental results on several datasets demonstrate that the improved method in this paper can highlight the detailed feature information in the image generation task and can effectively improve the quality of image generation.

    参考文献
    相似文献
    引证文献
引用本文

易星,潘昊,赵怀慈,刘鹏飞,杨斌.基于改进循环生成对抗网络实现红外图像生成[J].电子测量技术,2023,46(18):171-178

复制
分享
文章指标
  • 点击次数:619
  • 下载次数: 694
  • HTML阅读次数: 0
  • 引用次数: 0
历史
  • 在线发布日期: 2024-01-10
文章二维码
您是第5088791 位访问者
主管单位：北京电子控股有限责任公司
主办单位：北京无线电技术研究所
地址：北京市东城区北河沿大街79号
电话：13810562337
电子测量技术 ® 2025 版权所有