Scholarlink

<p>Remote sensing imagery plays an important role in the research and application of environmental change, such as phenology changes, land surface parameter relationships, land degradation, etc. However, remote sensing observation data are often missing due to sensor resolution limitations and environmental factors. The missing images may occur randomly or continuously, which is difficult to deal with this situation with conventional methods. To address this issue, we perform the multi-scene joint conditional generation network (JC-GN) based on radiation properties to learn the complex mapping between phase images. The proposed model introduces multi-scene joint conditional loss to constrain the spatio-temporal characteristics of the generated image. Our experiments, conducted utilizing unmanned aerial vehicle (UAV) datasets and the Landsat-8 datasets, which are located in the southeast of Gansu Province (34.73<InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="44443_2025_27_Article_IEq1.gif" Format="GIF" Height="7" Rendition="HTML" Resolution="72" Type="Linedraw" Width="9" /> </InlineMediaObject> <EquationSource Format="TEX">\(^{\circ }\)</EquationSource> <EquationSource Format="MATHML"><math> <mmultiscripts> <mrow /> <mrow /> <mo>∘</mo> </mmultiscripts> </math></EquationSource> </InlineEquation>N, 105.50<InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="44443_2025_27_Article_IEq1.gif" Format="GIF" Height="7" Rendition="HTML" Resolution="72" Type="Linedraw" Width="9" /> </InlineMediaObject> <EquationSource Format="TEX">\(^{\circ }\)</EquationSource> <EquationSource Format="MATHML"><math> <mmultiscripts> <mrow /> <mrow /> <mo>∘</mo> </mmultiscripts> </math></EquationSource> </InlineEquation>E), demonstrate that the proposed methodology achieves promising performance in terms of both precision (RMSE decreases by about 2.5 points, PSNR increases by about 10 points, and SSIM increases by about 0.03 points) and effectiveness (time cost savings of about 50 percent), and data-driven model exhibits a superior capacity to simulate intricate information with greater fidelity.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

JC-GN: multi-scene joint conditional network for remote sensing sequence imagery generation under spatio-temporal variation

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年4月10日

查看全文
  • Zi-yi Zhao,
  • Xing Jin

摘要

Remote sensing imagery plays an important role in the research and application of environmental change, such as phenology changes, land surface parameter relationships, land degradation, etc. However, remote sensing observation data are often missing due to sensor resolution limitations and environmental factors. The missing images may occur randomly or continuously, which is difficult to deal with this situation with conventional methods. To address this issue, we perform the multi-scene joint conditional generation network (JC-GN) based on radiation properties to learn the complex mapping between phase images. The proposed model introduces multi-scene joint conditional loss to constrain the spatio-temporal characteristics of the generated image. Our experiments, conducted utilizing unmanned aerial vehicle (UAV) datasets and the Landsat-8 datasets, which are located in the southeast of Gansu Province (34.73 \(^{\circ }\) ∘ N, 105.50 \(^{\circ }\) ∘ E), demonstrate that the proposed methodology achieves promising performance in terms of both precision (RMSE decreases by about 2.5 points, PSNR increases by about 10 points, and SSIM increases by about 0.03 points) and effectiveness (time cost savings of about 50 percent), and data-driven model exhibits a superior capacity to simulate intricate information with greater fidelity.