Skip to main content
. 2011 Oct 28;2(11):3207–3222. doi: 10.1364/BOE.2.003207

Algorithm 1.

Iteratively regularized Gauß-Newton algorithm

1: initialize
2: c ← 0
3: fork = 1 to maxit do
4:   [Ax, Am, G] ← assembleSystemMatrices(ck)
5:   Vx ← PBCG(Ax, Q)
6:   Vm ← PBCG(Am, GVx)
7:   MDVm
8:   Wm ← PBCG(Am, D)
9:   Wx ← PBCG(Ax, GWm)
10:   Sk ← assembleSensitivity(ck, Vx, Vm, Wx, Wm)
11:    dc(SkSk+αkI)1Sk(MδM)
12:   cc + dc
13:   αk+1k
14: end for