Efficient lq regularisation algorithm with range-azimuth decoupled for SAR imaging

The synthetic aperture radar (SAR) imaging methods based on compressive sensing endure great computational complexity since they have to deal with a large measurement matrix when processing the raw data. An efficient ℓ_q regularisation algorithm for SAR imaging is proposed, which uses a decoupling operator constructed by means of chirp scaling. It can reduce the computational complexity from the square order to the linear logarithm order compared with the conventional ℓ_q regularisation-based imaging algorithm while preserving the equivalent imaging quality. The proposed algorithm is efficient in the sense that it allows reconstruction of the sparse scenes with the sub-sampling data.