804 / 1363
Algebraic Reconstructions
Normally, a so-called „regularizer“ function is added to the cost function to be minimized,
in such a way to incorporate noise and/or artifacts reduction and create an algebraic
reconstruction algorithm optimized for a specific problem or scenario.
Here few examples:
- Total variation:
- Nuclear norm (for correlated images):
- Prior-induce similarity:
- Dicitionary based: