Scholarlink

In the field of remote sensing (RS), improving the spatial resolution of satellite imagery is critical for a variety of applications, including environmental monitoring, urban planning and forestry. With the recent advancements in deep learning, new opportunities have emerged for enhancing the spatial resolution of RS data. However, these models have been proved ineffective for tree detection tasks. In this paper, we present the GISSRGAN model, an extension of the ESRGAN architecture designed for single-image super-resolution (SISR) tasks involving satellite imagery. The model integrates auxiliary geospatial datasets, such as Landsat 7 ETM+ and Landsat 8–9 OLI/TIRS, to enhance super-resolution performance in forestry aimed tasks.

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

Super-Resolution of Satellite Images Using Landsat Data

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Ivan Sharshov,
  • Vladimir Berezovsky,
  • Kseniya Shoshina,
  • Roman Aleshko,
  • Irina Vasendina

摘要

In the field of remote sensing (RS), improving the spatial resolution of satellite imagery is critical for a variety of applications, including environmental monitoring, urban planning and forestry. With the recent advancements in deep learning, new opportunities have emerged for enhancing the spatial resolution of RS data. However, these models have been proved ineffective for tree detection tasks. In this paper, we present the GISSRGAN model, an extension of the ESRGAN architecture designed for single-image super-resolution (SISR) tasks involving satellite imagery. The model integrates auxiliary geospatial datasets, such as Landsat 7 ETM+ and Landsat 8–9 OLI/TIRS, to enhance super-resolution performance in forestry aimed tasks.