摘要:
A method for discriminating between a real face and a two-dimensional reproduction of the face, in a biometric detection process, the method comprising: (a) making at least two digital recordings of the face or its reproduction in time sequence, one after the other; (b) dividing each of the recordings into a number of image components wherein each image component comprises a number of pixels; (c) determining the displacement of the individual image components from the first recording to the second recording by correlation and generating a displacement vector field therefrom; and (d) analyzing the displacement vector field for determining whether the recording has been made from a real face or from its reproduction.