Through-the-wall target separation using low-rank and variational mode decomposition

Authors

Fok Hing Chi Tivive, University of WollongongFollow
Abdesselam Bouzerdoum, University of WollongongFollow

RIS ID

140488

Publication Details

F. Tivive & A. Bouzerdoum, "Through-the-wall target separation using low-rank and variational mode decomposition," IEEE Transactions on Geoscience and Remote Sensing, vol. 57, (12) pp. 9928-9940, 2019.

Abstract

1980-2012 IEEE. Through-the-wall radar imaging (TWRI) is a sensing technology that can be used for detecting, locating, and identifying targets inside enclosed building structures. Many of the existing target classification approaches focus on single-target scene. For multi-target classification, the radar signal has to be segregated into different target components. However, target separation in TWRI is a challenging problem since the radar signals consist of both strong wall reflections and weak target echoes. Furthermore, the target signals are attenuated and distorted when propagating through the wall. In this paper, a variational model with low-rank constraint is proposed for decomposing the radar signal into target components and removing the wall returns. Experimental results show that the proposed method can effectively separate the radar signal into different target components.