Volume 2 - No 5
FACIAL EXPRESSION DETECTION: A TECHNIQUE FOR OPTICAL FLOW ANALYSIS
This article describes a real-time automatic facial expression recognition system using a technique
for automatic facial expression detection based on Haartraning algorithm followed by automatic
points placement that performs facial movements’ reading using a technique for optical flow
analysis on these points. The facial movements read are interpreted as action units, according to a
classification presented by the researcher Ekman. As a result, an automatic real-time facial
expression and correlated emotion detection system capable of selecting one among six expressions
(anger, fear, happiness, sadness, disgust and surprise) was generated.
PaperID: p 472-489
Author's Name: Leonardo Panta Leão, Jonas Santos Bezerra, Leonardo Nogueira Matos and Maria Augusta Silveira Netto Nunes
Volume: Volume 2
Issues: No 5
Keywords: Image processing, Facial Expression, Emotion, Optical Flow, Action Units.