DEVICE AND METHOD FOR MULTI-VIEW FACE DATABASE, AND DEVICE AND METHOD FOR GENERATING MULTI-VIEW FACE DESCRIPTOR BY INTEGRATING FACE DESCRIPTORS OF MULTIPLE REGISTRATION VIEWS

摘要

PURPOSE: A device and a method for a multi-view face database, and the device and the method for generating a multi-view face descriptor by integrating the face descriptors of multiple registration views are provided to recognize a face of any pose and offer a high recognition performance by recognizing the face with the multi-view face descriptor. CONSTITUTION: An image obtainer(110) obtains multiple face images of diverse views including the face image of a front view from a user. An eye position determiner(130) selects multiple key frame images from the face images provided from the image obtainer and determines an eye position corresponding to each view included in a view mosaic by using the eye position of the selected key frame image. A face localizer(150) arranges a size and the eye position of the face image for each view by localizing the face image of each view included in the view mosaic based on the eye position of each view. A view mosaic generator(170) generates the view mosaic of the user by integrating the face image of each localized view.