Abstract
These days, the opportunities that social robots are active in various situations such like nursing, housework, and psychotherapy increase. The purpose of social robots is to fit into human everyday life and communicate with human. Now, most social robots use sound recognition or pressure recognition to have communication. However, it is difficult to read human emotions from only the information. If we are able to read expression, it makes possible to recognize more detail of human emotions because the part that human expresses their emotions mostly is facial expression. Then, we make a social robot that is able to read facial expression by image processing. In this paper, we show some steps to make the social robot. Firstly, it explains the development platform. Furthermore, we present the image processing to read facial expression and analysis of emotions. At the end, we show the conclusion and propose some improvement points.
