Typically, most research and academic institutions own and archive a great amount of objects and research related resources that have been produced, used and maintained over long periods of time by different types of "domain experts" (e.g. lecturers and researchers). Although the potential educational value of these resources is very high, this potential may largely be underused due to severe accessibility and manipulability constraints. The "virtualization" of these resources, i.e. their representation as reusable digital "learning objects" that can be integrated in an e-learning environment, would allow the full exploitation of all their educational potential. In this paper we describe the process model that we have followed during the virtualization of the objects and research resources owned by two academic museums at the "Complutense" University of Madrid (Spain). In the context of this model we also summarize the main aspects of these experiences in virtualization. (Contains 8 figures.)