Image analysis – Applications – Personnel identification
Reexamination Certificate
1999-01-14
2002-04-09
Boudreau, Leo (Department: 2621)
Image analysis
Applications
Personnel identification
C382S173000
Reexamination Certificate
active
06370263
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a method and device for registering and collating palm imprints of individual person who is to be identified for the personal authentication and, more particularly, to a method and a device capable of registering and collating palm imprints in a reduced time with simple means.
As is publicly known, fingerprint is the most popular physical information used in identifying the individual for his personal authentication. The fingerprints exhibit unique characteristics such as branching lines and end points. They differ among individual persons and do not vary throughout the life. Accordingly, disclosure has been made for a great number of devices and systems using the fingerprints or palm prints for the personal authentication. Furthermore, disclosure has been made for the devices utilizing palm prints to shorten the processing time and to improve the reliability.
For example, Japanese Unexamined Patent Publication (JP-A) No. 217984/1991, entitled “Fingerprint Collating device”, discloses a technique of shortening the fingerprint collating process. In this prior art, the device has the limited scanning range upon collating the fingerprint. Specifically, palm print images are preliminarily input into the collating device and then specified about which class the input palm print image belong to be classified. The fingerprint data are searched based on the palm print image class, and the input data and the database are sought for the matching data.
In the above-described fingerprint collating device, palm shapes are used in classifying the fingerprints to shorten the collating time used for authentication utilizing fingerprints. This device does not employ personal authentication making use of the palm prints so that the authentication accuracy is not good.
A technique for registering and collating palm prints is described, for example, in Japanese Unexamined Patent Publication (JP-A) No. 114640/1995, entitled “Personal Authentication System”. This authentication system extracts information characterizing the palm prints from a plurality of predetermined portions of each palm print and stores the extracted information. The collating process is carried out based on the stored data. The predetermined portions are determined based on the position of the middle finger and have characteristic palm print portions involving less positional errors. These predetermined portions are thinly extracted portions.
During authentication, features are extracted only from a few portions of the numerous portions described above. This improves the authentication speed. However, no mention is made about the means for selecting the above-mentioned few portions.
The aforementioned personal authentication system needs complex equipment including a prism, guide members, and a light source. These components are used to establish the positions where the few portions are extracted. That is, coordinates must be established in palm imprints finely and accurately.
Moreover, it is difficult to identify individual persons only with the few image parts because the same feature overlaps in many portions within the limited area which is thinly extracted. Consequently, a considerable number of image parts are necessary to obtain features from a palm print of the subject to be authenticated.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a method and device for identifying and collating individuals with palm prints easily and quickly with a reduced amount of data necessary for the collating.
In a method of registering palm imprints in accordance with the present invention, each palm imprint obtained from individuals to be identified is divided into plural parts or regions having predetermined positions and sizes when the palm imprints are registered for the authentication of the individuals. Feature points are extracted from the palm imprint parts and registered as information (hereinafter often referred to as “feature information”) about the feature points.
A device for registering a palm imprint in accordance with the present invention comprises an image display unit for displaying the palm imprint as image information on a display unit, a partial division unit for dividing the palm imprint into plural parts of regions having predetermined sizes and locations, a feature extraction unit for extracting feature points from the ranges of the image parts, and a data memory unit. Part information about the image parts and the feature information extracted by the feature extraction unit are correlated and stored in the data memory unit, together with corresponding information about the individual entered separately.
In this device, the palm imprint is divided into ranges having predetermined sizes and locations for authentication of individuals. This facilitates the discrimination and the identification of the individuals. The individuals can also be authenticated. In addition, the amount of data transmitted to and from the data memory unit can be reduced because the data is divided into parts.
In a method of collating a palm imprint in accordance with the invention, the feature points obtained from the palm imprint are registered corresponding to their respective image parts obtained by the registering method described above. When these feature points are compared against reference samples for authentication of the individual, he or she is authenticated if the feature information derived from at least one of the image parts makes a match.
A device for collating palm imprints of individuals in accordance with the invention comprises a data memory unit for storing part information about parts of each palm imprint and corresponding feature information about feature points as well as corresponding personal identity information for identification of the individuals, an image extraction unit for dividing each input palm print left behind and extracting at least one image part of the print, a feature extraction unit for extracting feature points from the image part or parts extracted by the image extraction unit, and an information collating unit for comparing the part information about the image parts and the feature information about the extracted feature points with data stored in the data memory unit and producing information that identifies verified individuals.
Where the collating is carried out by this device, feature information is extracted only from some image parts and used for the collating. This can reduce the amount of data transmitted to and from the data memory unit or the information collating unit. Moreover, the collating time can be shortened.
Other objects and features of the invention will appear in the course of the description thereof, which follows.
REFERENCES:
patent: 4186378 (1980-01-01), Moulton
patent: 4646352 (1987-02-01), Asai et al.
patent: 5067162 (1991-11-01), Driscoll, Jr. et al.
patent: 5073949 (1991-12-01), Takeda et al.
patent: 5526436 (1996-06-01), Sekiya
patent: 5528355 (1996-06-01), Maase et al.
patent: 52-142945 (1977-11-01), None
patent: 63-145529 (1988-06-01), None
patent: 3-71375 (1991-03-01), None
patent: 3-217984 (1991-09-01), None
patent: 7-21366 (1995-01-01), None
patent: 7-114640 (1995-05-01), None
patent: 8-129643 (1996-05-01), None
patent: WO 95/33242 (1995-12-01), None
Amanai Teruhito
Hiratsuka Seiichi
Yamada Yoshifumi
NEC Corporation
Werner Brian P.
LandOfFree
Method and device for registering and collating palm imprints does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and device for registering and collating palm imprints, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and device for registering and collating palm imprints will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2879550