Abstract:It is of great significance that a home service robot can recognize facial expressions of a human being. This thesis suggests that features of facial expressions be extracted with PCA, and facial expressions be recognized by distance-based Hashing K-nearest neighbor classification. First, Haar-like feature and AdaBoost algorithm is adopted to detect a face and preprocess the face image; then PCA is applied to extract features of the facial expression, those features will be inserted into the hash table; finally, the facial expression can be recognized by K-nearest neighbor classification algorithm. As concluded, recognition efficiency can be greatly improved after reconstructing the feature database into hash tables.

Facial Expression Recognition of Home Service Robots