DEVICE AND METHOD FOR FACE FEATURES SELECTION USING BOOSTING ALGORITHM AND FEATURE VECTOR SEGMENTATION

摘要

PURPOSE: A face feature selection method using boosting algorithm and feature vector segmentation and a device thereof are provided to drastically reduce time spent for face feature selection using boosting algorithm and feature vector segmentation, thereby improving processing speed of a face identifying system. CONSTITUTION: A feature vector group is inputted and saved(S110). A learning weighted value is given to each feature vector of the feature vector group(S120). The feature vector group Is partitioned into one or more sub vector groups(S130). A group in which learning errors are minimum is extracted. A face feature is selected from the extracted sub vector group.