LAUSR

Inverse synthetic aperture radar (ISAR) image data of a single target is sparse in the image domain. Based on this sparseness, we could obtain a high-precision image reconstruction by down sampling the imaging data and getting the sparse solution of the indeterminate equations. In this work, we have studied the sparse data processing theory based on the compressed sensing (CS) method. We focus on the sparse reconstruction of the ISAR image. The imaging data are sparsely sampled and restored through the norm regularization framework. We compare the reconstruction results on $L_{1}$ and $L_{\mathrm {1/2}}$ regularization frameworks, respectively. Then, we concentrate on the relationship between the reconstruction results and parameter settings in the reconstruction framework. Besides, we study the ISAR image in different radar bands. The numerical results show that the $L_{\mathrm {1/2}}$ regularization framework is better than the $L_{1}$ framework in recovery accuracy and computational efficiency.