Facial expression recognition method and system based on motor unit layering

Abstract

A method and system for facial expression recognition based on motor unit layering. The invention discloses a facial expression recognition method, which includes three-layer classification steps, specifically: firstly, extracting the area adjacent to the upper part of the nose as the first-layer classification area to detect whether the AU9 motion unit appears as the first-layer classifier judgment standard , to roughly classify the expression; then, extract the lip area as the second-level classification area, to detect whether the AU25 and AU12 motor units appear as the second-level classifier criteria to subdivide the first-level classification results; finally , respectively extract the upper and lower half face regions as the third-level classification region for fine classification in the second-level classification results. The present invention also provides a system for realizing the above method. Based on the AU layered structure, the present invention extracts the features of representative regions for expressions and classifies them layer by layer with random forests, effectively improving the accuracy and speed of expression recognition, and is especially suitable for low-resolution images.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a method and system for recognizing facial expressions. Background technique [0002] Facial expression recognition is to classify facial expressions by analyzing facial movements and changes in facial features through visual signals. The research on expression classification is basically based on the six main human emotions first proposed by psychologists Ekman and Friesen in 1971. Each emotion reflects a unique psychological activity of a person with a unique expression. These six emotions are called basic emotions and consist of anger, happiness, sadness, surprise, disgust and fear. In recent years, with the rapid development of a series of related fields, such as machine learning, image processing, face detection, etc., more and more attention has been paid to facial expression recognition, and its application prospects are very broad, such as natural human-computer ...

Images