Remote Sensing (Nov 2023)

The LPR Instantaneous Centroid Frequency Attribute Based on the 1D Higher-Order Differential Energy Operator

  • Xuebing Zhang,
  • Zhengchun Song,
  • Bonan Li,
  • Xuan Feng,
  • Jiangang Zhou,
  • Yipeng Yu,
  • Xin Hu

DOI
https://doi.org/10.3390/rs15225305
Journal volume & issue
Vol. 15, no. 22
p. 5305

Abstract

Read online

In ground-penetrating radar (GPR) or lunar-penetrating radar (LPR) interpretation, instantaneous attributes (e.g., instantaneous energy and instantaneous frequency) are often utilized for attribute analysis, and they can also be integrated into a new attribute, i.e., the instantaneous centroid frequency. Traditionally, the estimation of instantaneous attributes calls for complex trace analysis or energy operator schemes (e.g., the Teager–Kaiser energy operator, TKEO). In this work, we introduce the 1D higher-order differential energy operator (1D-HODEO) to track instantaneous attributes with better localization. In collocation with the mode decomposition algorithms, the 1D-HODEO performs along each A-scan on the decomposed mode slices to form the final profile of instantaneous centroid frequency by using the piece-wise correlation coefficients. Both a numerical model for simulating two-layer lunar regolith and the LPR Yutu-2 data show that the proposed instantaneous centroid frequency profile on the 1D-HODEO has better resolution, in comparison with that of TKEO and the traditional time-varying centroid frequency. In this work, we present a new approach for extracting instantaneous centroid frequency attributes which provides more comprehensive information in lunar stratigraphic interpretation and LPR attribute analysis.

Keywords