Post-processing - binarization with matlab running under a batch script.
1. Set the jobs/paths
    a. Open ./batchBin
    b. Edit the comment/uncomment sections; change the exp number as needed
    c. Set where the .tif stacks live. This should be an easily reconstructed path 
       (e.g. differs only in exp numbers). The Raw_data folder can be deconstructed for this
       purpose, the file names in that folder will never conflict; file names in the
       "reconstructions" folder will not conflict with each other, although two copies of
       each exp number exist in the full raw dataset (from June and November test events),
       so these need to be deconflicted somehow, e.g. a "nov" and "jun" higher-level folder.
    d. Set the output directory; a sub-directory "matBin" will be generated if it is 
       not already there.
2. Set up matlab options
       Each scan/material/setup might require the different smoothing, bwopen, and 
       thresholding options to be tuned by hand a little. Use "testFiltering.m" to 
       get something that looks good for the whole stack (test a few images) and put 
       those options into the "tif2bin.m" script, replacing the current filtering etc. 
       options (though backing up the old one is a good idea). The matlab command 
       imshowpair requires a figure window, so running this step with graphics on a 
       local computer is generally advised.
3. Executing $ ./batchBin will automatically generate and submit jobs for each experiment
   requested, if operating on a MAUI-based HPC scheduler. Each binarization runs on 24
   processors and should take about 19-25 minutes on processors working at roughly 4 GHz. Job
   submission scripts may need to be adapted/rewritten for your computational environment.
   A stack of binarized tif images will be produced along with at matlab .mat file that contains
   a 3D binary array that is equivalent to the tif stack. A scaled-down .mat file will also be 
   generated.