A facial expression recognition system using neural networks

摘要: This paper proposes an automatic facial expression recognition system using neural network classifiers. First, we use the rough contour estimation routine, mathematical morphology, and point detection method to extract precise contours of eyebrows, eyes, mouth a face image. Then define 30 characteristic points describe position shape these three organs. Facial expressions can be described by combining different action units that are used for describing basic muscle movement human face. We choose six main units, being composed movements, as input vectors two network-based classifiers including radial basis function multilayer perceptron network. Using networks, have obtained same rate high 92.1%. Simulation results computer demonstrate computers capable extracting high-level or abstract information like human.