Timeline for A notion of resolution in inverse problems
Current License: CC BY-SA 4.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 21, 2020 at 22:45 | vote | accept | Ron | ||
| Dec 21, 2020 at 20:36 | comment | added | Wolfgang Bangerth | @Ron That depends on what you want to test and what the application is. If you have a deblurring problem, then you want to know how far a one-pixel noise in the original image $b$ spreads to the deblurred image $x$. On the other hand, if your $A$ is modeling a forward process, then you will want to know how sharp a reconstruction of a point source would be, and in that case the perturbation $\delta b$ should be $Ae_i$. | |
| Dec 20, 2020 at 16:43 | comment | added | Ron | Shouldn't the change $e_i$ be given by feeding a unit change $c_i$ into the forward operator, i.e. $e_i=Ac_i$? Because $B$ denotes only $A^\dagger$ in this case. | |
| Dec 18, 2020 at 22:52 | history | answered | Wolfgang Bangerth | CC BY-SA 4.0 |