About vxl/vnl/vcl math functions (refactor)
As part of code review of !162 we identified that functions such as `vcl_abs` are not really need anymore. There are [defined to their std equivalents](https://github.com/InsightSoftwareConsortium/ITK/blob/master/Modules/ThirdParty/VNL/src/vxl/vcl/vcl_compiler.h#L339) in ITK master branch.
See also ITK commit [21334467](https://github.com/InsightSoftwareConsortium/ITK/commit/2133446772e6a688e25cbfb1b74bb709103fe77f):
> ENH: Avoid using vnl_math_ functions
>
> In order to isolate ITK from vnl_math_ functions an aliasing
> system is created to prefer to use itk::Math:: functions
> over the vnl_math_ versions. This is partially precipated by
> a need to provide backwards compatibility with vnl_math_ prior
> to the introduction of the vnl_math:: namespace so that
> we can consistently use it.
>
> Recent cleanup of the vxl codebase identified that modern
> supported compilers do not need most of the aliasing infrastructure
> of vxl/vnl/vcl and that in many cases functions simply degrade to
> the std:: versions. In order to minimize dependancies, and to
> assist new users of ITK to more naturally understand the code,
> we prefer to use the std:: versions whenever possible.
>
> In particular, there is no benefit to using vnl_math_min or
> vnl_math_max functions that always degenerated to equivalent
> functionality as std::min and std::max. Since the std:: versions
> are more universally understood and known by most developers,
> we prefer to use these common versions.
It seems like we could use `std::` functions most of the time?
issue