Method for distinguishing real face and two-dimensional face image in the process of biological data collection

Abstract

The present invention relates to a method of distinguishing between a real face and a two-dimensional image of a face in the collection of biological data, wherein (a) at least two digital records of the face or its image are sequentially captured in chronological order; (b) Each record is decomposed into several image units, and each image unit includes several pixels; (c) determining the displacement of each image unit between the first record and the second record through a correlation algorithm, and generating a displacement vector field accordingly; and ( d) Analyzing the displacement vector field to determine whether a real face or its image is captured. Furthermore the invention relates to a method of identifying a person in a distributed IT infrastructure, especially in a cloud environment, wherein at least two digital photographic records of the person's face are received on a computer remote from the person, and then Distributed execution of the method of any one of the preceding claims on one or more computers remote from the person, wherein the person's facial record is also used for identification.

Description

technical field [0001] The present invention relates to a method for distinguishing a real face and a two-dimensional face image during biological data collection, and a method for identifying a person using the distinguishing method. Background technique [0002] Biometric identification based on facial recognition has long been known. For example, the digital record of the face can be compared with a reference photograph of the face, which was taken for example according to the ICAO (Internationale Zivilluftfahrt-Organisation, International Civil Aviation Organization) regulations in accordance with the biometric standard ICAO 9303 (Photograph Guideline). Comparing digital records and photographs is routine at many border checkpoints. It is also possible to compare a person's digital record with the person's record held in a database in order to allow the person to access devices, computers, Internet applications, and the like. For example, German patent DE 19847261 desc...

AI Technical Summary

Problems solved by technology

[0014] Although the optical flow method is theoretically suitable for distinguishing between photographic records and photographic images of real human bodies, it faces a series of problems in practice: calculating optical flow pixel by pixel requires a lot of computing resources, which makes it difficult to check within a reasonable time

Due to the high noise content of the digitally recorded pixels, the resulting flow vector needs to be smoothed through several images, again increasing the data volume and computational expense

Optical flow is also error-prone

Images