The focus of this Diploma thesis is the development of a new approach for the analysis of crystal orientations in polar ice, based on digital image processing. The c-axes orientations are measured on thin ice sections using automated polarized light microscopy. The measured data are stored in form of high-resolution digital images which have been evaluated only manually so far. The automatic detection of grains and grain boundaries and the extraction of their orientation and shape parameters enable a statistical acquisition of the ice microstructure. The analysis of c-axes distributions, grain sizes, misorientations between adjacent grains, grain boundary orientations etc. allows for conclusions about the active recrystallization mechanisms. This is of high importance for an improved understanding of ice deformation and ice flow. The developed program is applied to analyze the microstructure in the NEEM ice core (Greenland).