Recent technology trends in IT are shifting from device-centered to user-centered. Furthermore, human vision plays a significant role to recognize objects and situations, and computers and robots need methods to evaluate and analyze complex human emotions. Yet, one of the serious and well-known social issues is that although lots of modern people feel stressed and psychological anxiety in a workplace, home, etc., they are afraid of expressing their emotions. Hiding emotions or failing to release stress can cause more serious medical issues. For this reason, this research focused on developing the system of recognizing facial expressions to provide psychological diagnosis and treatment. In this research, we developed a facial recognition system for psychotherapy robots. The YCbCr color model was used for image pre-processing, and face detection was performed by the Haar-like feature. For the facial images detected, principal components analysis (PCA) was used to recognize the expressions of the seven basic emotions: expressionless, happy, sad, fearful, disgusted, surprised, and angry. According to the categorized facial emotions, the picture related to psychotherapy was successfully presented.
Volume 11 | 12-Special Issue
Pages: 953-958
DOI: 10.5373/JARDCS/V11SP12/20193299