I you want to avoid using manual evaluation or image processing, you should use SALS (small angle light scattering) from which you can calculate directly the average spherulite size.

I am puzzled by the question.  Polymer crystallites are tiny and not individually visible in an optical microscope.  You can use small angle X-ray scattering, or possibly Raman.  As Gyorgy says, if it is spherulites you are looking for, they may be visible, depending on the polymer and nucleation density etc, but small angle light scattering would be a suitable tool for this.

this article may be useful.

http://pubs.rsc.org/en/content/articlelanding/2013/sm/c2sm26988k#!divAbstract

Thank a  lot for your answers. But not that what I looking.