A Gaussian-Rayleigh mixture modeling approach for through-the-wall radar image segmentation

Authors

Cher Hau Seng, University of WollongongFollow
Abdesselam Bouzerdoum, University of WollongongFollow
Moeness G. Amin, Villanova UniversityFollow
F Ahmad, Villanova University

RIS ID

69526

Publication Details

C. Seng, A. Bouzerdoum, M. G. Amin & F. Ahmad, "A Gaussian-Rayleigh mixture modeling approach for through-the-wall radar image segmentation," in ICASSP 2012: IEEE International Conference on Acoustics, Speech and Signal Processing, 2012, pp. 877-880.

Abstract

In this paper, we propose a Gausssian-Rayleigh mixture modeling approach to segment indoor radar images in urban sensing applications. The performance of the proposed method is evaluated on real 2D polarimetric data. Experimental results show that the proposed method enhances image quality by distinguishing between target and clutter regions. The proposed method is also compared to an existing Neyman-Pearson (NP) target detector that has been recently devised for through-the-wall radar imaging. Performance evaluation of both methods shows that the proposed method outperforms the NP detector in enhancing the input images.