DEM generation using bistatic interferometry: High-coherence pixel selection and residual reference phase compensation

Tao Zeng, Mao Zhu, Cheng Hu, Weiming Tian, Michail Antoniou

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)

Abstract

The bistatic spaceborne/stationary InSAR configuration could be applied to generate digital elevation model (DEM). However, two factors limit the DEM precision improvement. On the one hand, due to the effects of the illuminator’s sidelobe and the high objects’ shadow, the pixels in the corresponding area might be polluted by the noise. On the other hand, owing to the presence of deviation (between the real value and the nominal value of the master/slave antenna positions), after utilizing the nominal value to compensate for the reference phase, the residual reference phase would remain. In order to solve the problems, we proposed a new bistatic DEM generation algorithm. Firstly, we applied the coherence-phase joint analysis method to select high-coherence points. Secondly, we deduced the transforming relationship from the deviation to the residual reference phase, and then compensated for the residual reference phase. In the experiment, the YaoGan-3 (an L-band spaceborne SAR) was selected as the transmitter and the two stationary receivers were mounted on the top of a tall building. The algorithm recommended in this paper was applied to process the measured bistatic data. The analysis of the results demonstrated that the elevation error of the generated DEM for some targets had been limited to the level of 1–2 m.

Original languageEnglish
Pages (from-to)1-14
JournalScience China. Information Sciences
Volume58
Issue number6
Early online date30 Apr 2015
DOIs
Publication statusPublished - Jun 2015

Keywords

  • bistatic InSAR
  • bistatic SAR
  • DEM
  • high coherence pixel
  • SAR

ASJC Scopus subject areas

  • Computer Science(all)

Fingerprint

Dive into the research topics of 'DEM generation using bistatic interferometry: High-coherence pixel selection and residual reference phase compensation'. Together they form a unique fingerprint.

Cite this