. 2015 Feb 12;15(2):4176–4192. doi: 10.3390/s150204176

Table 1.

The proposed ASR-CS-SAR approach.

1.	Initialization:
	(a) Compute an initial image q⁽⁰⁾ by solving Equation (10) with the wavelet basis selected;
	(b) Suppose l = 1, ω⁽⁰⁾ = 0, λ⁽⁰⁾ = 1.5, δ_e = 10⁻⁶, and let I_max be the maximum iterations;
2.	Updating the parameter matrix D⁽^l⁾
	(a) Compute the SSR of the image according to Equation (12);
	(b) Update the matrix D⁽^l⁾ according to Equations (14) and (16);
3.	SAR imaging with fixed A⁽^l⁾ = I − D⁽^l⁾
	(a) Update ω⁽^l⁾ according to Equation (22) with fixed q⁽^l⁾;
	(b) Update q⁽^l⁾ according to Equation (23) with fixed ω⁽^l⁾;
	(c) Update the related parameters according to Equation (24);
4.	If, ${{‖ q^{(l)} - q^{(l - 1)} ‖}_{2}^{2} / ‖ q^{(l - 1)} ‖}_{2}^{2} < δ_{e}$ or l ≥ I_max output the image q⁽^l⁾; otherwise let l = l + 1 and jump to 2.