Baseline distribution optimization and missing data completion in wavelet-based CS-TomoSAR

Abstract

In this paper, we propose a coherence of measurement matrix-based baseline distribution optimization criterion, together with an L ₁ regularization missing data completion method for unobserved baselines (not belonging to the actual baseline distribution), to facilitate wavelet-based compressive sensingtomographic synthetic aperture radar imaging (CS-TomoSAR) in forested areas. Using M actual baselines, we first estimate the optimal baseline distribution with N baselines (N > M), including N − M unobserved baselines, via the proposed coherence criterion. We then use the geometric relationship between the actual and unobserved baseline distributions to reconstruct the transformation matrix by solving an L ₁ regularization problem, and calculate the unobserved baseline data using the measurements of actual baselines and the estimated transformation matrix. Finally, we exploit the wavelet-based CS technique to reconstruct the elevation via the completed data of N baselines. Compared to results obtained using only the data of actual baselines, the recovered image based on the dataset obtained by our proposed method shows higher elevation recovery accuracy and better super-resolution ability. Experimental results based on simulated and real data validated the effectiveness of the proposed method.