Image analysis – Applications – Personnel identification
Patent
1996-11-26
2000-08-08
Couso, Jose L.
Image analysis
Applications
Personnel identification
382155, G06K 900, G06K 962
Patent
active
061012645
DESCRIPTION:
BRIEF SUMMARY
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to the identification of individuals by evaluation of information inherent to the individual itself and without using auxiliary means, such as identity paper, keys or personal identification numbers
2. Prior Art
A reliable recognition--thus an identification--of individuals still remains technically unsolved. Current approaches are based on individuals being identified by auxiliary means (examples for this are the often falsified identity paper, the lost key or the too often forgotten personal identification number). Nowadays, also passwords have an only short duration, since they either have a too simple form or may easily be found out by unauthorized persons. It is just the last mentioned danger which frequently occurs when an authorization of access is urgently needed and the entered password (supposed to be correct) does not work.
Reliable identification of individuals, however, also means: protection of a local area against unauthorized access. Examples for this are military installations or danger zones in nuclear power plants. Also, the access to information, such as files or secret service information, which nowadays become more and more important, requires an exact coordination of the authorization of access for persons, dependent on their reliability and competence.
From prior art, the analysis of static images is known, such as face recognition, recognition of handprints or analysis of fingerprints. Due to great variations of an individual's outer appearance and also due to the difficulty in providing reproducible recording conditions, said systems show an "inaccuracy" which is still so high that one cannot speak of a "reliable identification of individuals". The identification of individuals can only be regarded as reliable when working with (almost) 100% accuracy.
SUMMARY OF THE INVENTION
The invention proposes a process for the identification of individuals (humans), wherein movement information from image sequences (FIG. 2) originating from a typical or predetermined movement action at or of said individual, is recorded, evaluated and compared with comparative values with the invention proposes a synergetic computer for the realization of the aforementioned processes.
Instead of the static image according to prior art, the invention uses image sequences. Said image sequences define the movement information being evaluated by calculating an optical flow (compare FIG. 6 and FIG. 7). In order to provide reliable evaluation, at least one of a typical and a predetermined movement action is used, the image sequences of which are evaluated and compared with comparative values. Before carrying out such comparison, the data may be reduced by a classification process, said data reduction providing a filter for large data quantities.
A typical movement action may be at least one of a walk and a head movement; a predetermined movement action may be the movement of a face or a section of a face (lineaments), which movement results from speaking a predetermined word or sentence.
Advantageously, said typical and predetermined movement actions may be combined. The evaluation of said combination may be effected in parallel or interleaved.
An evaluation of static image information may be used in addition to said image sequence information.
Said image sequence information and the evaluation of said static image information may be supplemented by an evaluation of acoustic information to (further) increase reliability.
The subject matter of the invention is directed to a synergetic computer for the realization of the identification process in the aforementioned sense; besides, it is described more in detail in a conference report volume of a conference in Innsbruck (April 1993; pages 206 to 212). The title of said document is "Using a Synergetic Computer in an Industrial Classification Problem"; the subject matter of said conference and the name of said volume is "Artificial Neural Networks and Genetic Algorithms". Particularly item
REFERENCES:
patent: 4757541 (1988-07-01), Beadles
patent: 4841575 (1989-06-01), Welsh et al.
patent: 4975960 (1990-12-01), Petajan
patent: 5161204 (1992-11-01), Hutcheson et al.
patent: 5163111 (1992-11-01), Baji et al.
patent: 5164992 (1992-11-01), Turk et al.
patent: 5274714 (1993-12-01), Hutcheson et al.
patent: 5410609 (1995-04-01), Kado et al.
patent: 5412738 (1995-05-01), Brunelli et al.
patent: 5426745 (1995-06-01), Baji et al.
patent: 5432864 (1995-07-01), Lu et al.
patent: 5497430 (1996-03-01), Sadovnik et al.
patent: 5586215 (1996-12-01), Stork et al.
patent: 5621858 (1997-04-01), Stork et al.
Wagner et al. (Multi-Sensorial Inputs for The Identification of Persons W Synergetic Computers, IEEE, pp. 287-291), 1994.
Mase et al., Automatic Lipreading by Optical-flow Analysis, System and Computer in Japan, vol. 22, No. 6, 1991.
Haken, Synergetic, IEEE Circuits and Devices Magazine, Nov. 1988.
Zakhor et al., Edge-Based 3-D Camera Motion Estimation With Application to Video Coding, IEEE Transactions on Image Processing vol. 2, No. 4 (1993), New York, US, pp. 481-498.
Hartwig et al., Digitale Bildcodierung, (Teil 12) FKT Fernseh-und Kino-Technik vol. 47, No. 1, (1993) Heidelberg, DE, pp. 33-42.
Wagner, et al., Using a Synergetic Computer in an Industrial Classification Problem, Artificial Neural Networks and Genetic Algorithms, (Apr. 1993) Innsbruck, AT, pp. 206-212.
Yamasaki et al., Advanced Intelligent Sensing System Using Sensor Fusion, IECON'92, (1992) San Diego, CA, US, pp. 1-8.
Kenji, Automatic Lipreading by Optical-flow Analysis, Systems and Computers in Japan, vol. 22, No. 6, (1992) Japan, pp. 67-76.
Horn et al., Determining Optical Flow, Artificial Intelligence Laboratory, Massachusetts Institute of Technology, (1981) Cambridge, MA, US pp. 185-203.
Bauer Norbert
Boebel Friedrich G.
Wagner Thomas
Couso Jose L.
Dang Duy M.
Fraunhofer Gesellschaft Fuer Angewandte Forschung e.V. et al
LandOfFree
Person identification based on movement information does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Person identification based on movement information, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Person identification based on movement information will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-1157731