IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (Jan 2020)

Clutter Removal of Near-Field UWB SAR Imaging for Pipeline Penetrating Radar

  • Alireza Akbarpour,
  • Somayyeh Chamaani,
  • Juergen Sachs,
  • Giovanni Del Galdo

DOI
https://doi.org/10.1109/JSTARS.2020.2983891
Journal volume & issue
Vol. 13
pp. 1527 – 1539

Abstract

Read online

Recently, ultrawideband (UWB) near-field synthetic aperture radar (SAR) imaging has been proposed for pipeline penetrating radar applications thanks to its capability in providing suitable resolution and penetration depth. Because of geometrical restrictions, there are many complicated sources of clutter in the pipe. However, this issue has not been investigated yet. In this article, we investigate some well-known clutter removal algorithms using full-wave simulated data and compare their results considering image quality, signal to clutter ratio and contrast. Among candidate algorithms, two-dimensional singular spectrum analysis (2-D SSA) shows a good potential to improve the signal to clutter ratio. However, basic 2-D SSA produces some artifacts in the image. Therefore, to mitigate this issue, we propose “modified 2-D SSA.” After developing the suitable clutter removal algorithm, we propose a complete algorithm chain for pipeline imaging. An UWB near-field SAR monitoring system including an UWB M-sequence sensor and automatic positioner are implemented and the image of drilled perforations in a concrete pipe mimicking oil well structure as a case study is reconstructed to test the proposed algorithm. Compared to the literature, a comprehensive near-field SAR imaging algorithm including new clutter removal is proposed and its performance is verified by obtaining high-quality images in experimental results.

Keywords