Scholarlink

<p>Interferometric Synthetic Aperture Radar (InSAR) has been widely used in terrain mapping due to its high spatial resolution and all-weather capability. However, the presence of significant noise and redundant information in interferograms often limits the accuracy of Digital Elevation Model (DEM) reconstruction. This paper proposes an improved Maximum A Posteriori (MAP) estimation method enhanced with Principal Component Analysis (PCA). By extracting the dominant features from high-dimensional interferometric data, the approach effectively suppresses noise and reduces dimensionality, thus improving computational efficiency while maintaining high reconstruction accuracy. Experimental results demonstrate that the proposed method achieves comparable accuracy to conventional MAP algorithms, with a nearly 20-fold increase in processing speed, making it highly suitable for large-scale data applications.</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

Toward efficient and accurate InSAR elevation modeling via Dimensionality-Reduced MAP estimation

  • 期刊论文

  • 发表日期：

    2026年4月23日

查看全文
  • Wei-jian Liu,
  • Zhuang-zhuang Wang,
  • Zi-wei Li,
  • Zhi-zeng Zhang,
  • Zhong-kai Peng,
  • Zhen-xia Yuan,
  • Xin-bo Luan

摘要

Interferometric Synthetic Aperture Radar (InSAR) has been widely used in terrain mapping due to its high spatial resolution and all-weather capability. However, the presence of significant noise and redundant information in interferograms often limits the accuracy of Digital Elevation Model (DEM) reconstruction. This paper proposes an improved Maximum A Posteriori (MAP) estimation method enhanced with Principal Component Analysis (PCA). By extracting the dominant features from high-dimensional interferometric data, the approach effectively suppresses noise and reduces dimensionality, thus improving computational efficiency while maintaining high reconstruction accuracy. Experimental results demonstrate that the proposed method achieves comparable accuracy to conventional MAP algorithms, with a nearly 20-fold increase in processing speed, making it highly suitable for large-scale data applications.