Add `postcorrect` notebook

[arentkievits]

We can use the postcorrector.py code in the scripted_render_pipeline repo to do interactive optimization of the post-correction parameters (artifact detection). Based on existing notebook from Ryan. The script can then use the optimized post-correction parameters for the full stack.

• [x] Import functions into notebook
• [x] Call functions for post-correction, do not write new tiffs to disk
• [x] Visualize subset of data (i.e. one or two sections) using standard parameters
• [x] Easy way to change the parameters and see the result on a small data set.
• [x] (Optionall) call post_correct script directly from notebook with optimal parameters

[arentkievits]

This is relevant because some datasets have ROAs that partly cover the scintillator surface. This results in gradients in all tiles, because these tiles are somehow not detected as outliers.