Resumen: The aim of this Bachelor’s Thesis is to provide the therapeutic robot TUK with an identification software, so that it is able to recognize people in its environment. Moreover, it is intended that TUK can identify who it is interacting with, and thus adapt its behaviour depending on the situation. In the first part, the theoretical principles that have been used during the development of the software are presented. Several kinds of image processing techniques as well as classification algorithms are explained. In the second part, the implementation is shown step by step in order to give an overview of the whole system. Finally, the results obtained during several tests are presented and discussed. In conclusion, several guidelines for tackling some of the challenges are proposed, setting a possible way for further work.