WIP!: Img seg

[giltis]

This is the running github issue for the image segmentation example for micro-ct glass bead packed column. This example demonstrates how to extract bead volume, pore space and calculated porosity of the packed column.

There is an ipython notebook and a VisTrails example. Both of these are included with this PR and will be updated as new functionality is added in PRs against scikit-xray

IPython notebook link: http://nbviewer.ipython.org/gist/anonymous/529b56087698dd0f9569
(Note that this is an old segmentation workflow and will be updated as the tools get implemented in scikit-xray.)
(Also note that this notebook needs to be edited to work again...)

This example will demonstrate the following functionality:

Step 1

• Load image (Note that this will likely need to be special-cased for different types of input data)
• View Raw Image (Using matplotlib)
• Image subtraction (Image processing: image arithmetic and logic tools scikit-beam#260)
• Plot Subtracted Images (Using matplotlib)

Step 2

• Image alignment
• Plot aligned images (Using matplotlib)
• Image alignment (above the absorption edge and below the absorption were collected at different times)

Step 3

• Image Filtering (make sure you are plotting what is going to be thresholded)
• Compute histogram
• Plot histogram (Using matplotlib)

Step 4

• Threshold by providing a specific value
• Automatic thresholding
  • Otsu
  • Yen
  • Adaptive
  • Iterative

Step 5

• Isolate pore space

Step 6

• Generate mask that represents the walls of the packed column
• Plot!

Step 7

• Create a label field that distinguishes the beads from the container and the empty volume
• Plot to verify that the label field is correct
• Quantify the volume and porosity

Step 8

• Render the label field in 3-D