Implementation

Goal: Maximize

By adjusting both the Image and the Model.

In practice, find the image-model pair with the fewest Pixons,

each with as many counts a possible, while still satisfying the

constraints given by the data.