For an overview of the topic, please refer to the pmbfilter page (this function is used as part of this method).

 VMR::InhomogeneityCorrect  - attempt automatic inhomogeneity correction
 
 FORMAT:       [vmr = ] vmr.InhomogeneityCorrect([opts])
 
 Input fields:
 
       opts        optional struct with settings
        .mask      either 3D uint8/logical data or VMR object with preseg
                   if omitted, try automatic mask detection
        .model     either of 'log', {'mult'}
        .numpasses number of passes, default 3 (valid: 1 through 5)
        .order     polynomial order, default 3 (valid: 2 through 7)
 
 Note: this function uses pmbfilter.

The .mask, .model, and .order options are passed through to pmbfilter.

The .numpasses option refers to how many times pmbfilter is called. On the last iteration, the call uses the .cmask and .robust options of pmbfilter (set to true) for an improved final fit.

If the VMR object contains valid 16-bit data (V16 file), the filtering will be applying to this data instead, followed by a call to VMR::LimitVMR with option .recalc8b set to true.