US20020174347A1 - Authentication with variable biometric templates - Google Patents

Abstract

The invention relates to systems and methods for using a template in the authentication process using biometric data. In one embodiment, a module modifies a template of the reference set of biometric data with the candidate set of biometric data when the user is authenticated. In another embodiment, a module modifies a copy of the template of the reference biometric data with modification data thereby creating a challenge template. The client compares the challenge template to a candidate set of biometric data thereby creating a response vector. A module authenticates the user based on the response vector and the modification data.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS
• This application claims the benefit of and priority to the co-pending U.S. Provisional Application Serial No. 60/291,900, filed May 18, 2001, entitled “Network-Based Biometric Authentication,” the entirety of which is incorporated herein by reference.[0001]
FIELD OF INVENTION
• The invention relates generally to biometrics. More specifically, in one embodiment, the invention relates to systems and methods for using biometric authentication over a network.[0002]
BACKGROUND
• The Internet accords a global community of computer users access to applications and information that traditionally were highly restricted. For example, users can now undertake a wide variety of financial transactions online, or obtain access to financial and other sensitive records online. The increased accessibility of such information, while enormously convenient, jeopardizes privacy and invites tampering and electronic theft. In some known prior art systems, sensitive information that was once physically guarded can now be obtained on the Internet by anyone who can generate the correct server URL, logon and password.[0003]
• Indeed, the mere need for Internet users to keep track of multiple URLs, logon names, passwords and PINs in order to access different information further increases the chances of unauthorized use and loss of private information. Users may resort to using the same logon name and password combinations for all accounts, rendering them equally vulnerable if unauthorized access to a single account is obtained. On the other hand, security-conscious users who maintain different logon names and passwords for individual accounts may, to avoid confusion, write them down where they may be found or store them on easily stolen devices such as personal digital assistants—thereby undermining their own efforts. It can be argued that those who routinely change their passwords but record them on paper or in a computer file are at greater risk of being compromised than those who use a single but difficult-to-crack password. At the very least, such security-conscious individuals risk forgetting their access information, necessitating time-consuming calls to customer-support lines.[0004]
• From the perspective of authentication, passwords and PINs cannot guarantee identity; the identification is no more reliable than the security of the password. In some known prior art systems with password authentication, the server carrying out a transaction can only prove that the correct password was entered—not that it was entered by an authorized person. A password can originate from password-cracking software just as easily as from the real user. Digital certificates improve security by authenticating an end point (i.e., that a message originated with a particular client terminal), but cannot create a non-repudiated link to support the claim that a particular user really did engage in a transaction.[0005]
SUMMARY OF THE INVENTION
• The present invention utilizes biometric indicia to offer highly reliable authentication that creates links that cannot be repudiated for transactions initiated within the context of an authenticated session. Unlike passwords, which are no more than secrets vulnerable to theft, biometrics validation matches physical characteristics of the user against stored characteristics to identify the user. Once a user is positively identified, in one embodiment, the server unlocks and validates the user's credentials for presentation to other servers that request such authentication. A user's credentials may, for example, represent an account login/password combination or X.509 certificate. This biometric approach offers substantial flexibility in terms of accessibility (from computers, mobile devices, etc.) and relieves the user from responsibility for managing the integrity of such credentials. Biometric scanners are inexpensive and small, and may, for example, be easily incorporated into keyboards and mobile client devices.[0006]
• In one embodiment, the authentication process can use an adaptive learning algorithm to improve the accuracy and reliability of matching a candidate set of biometric data against a user's biometrics profile (e.g., a reference set of biometric data stored as, for example, a template). Candidate sets of biometric data that result in successful matches are used to augment the profile and improve the statistics need to establish a subsequent reliable match. Upon authentication, new biometric data are introduced into the reference set associated with the profile if it is dissimilar or covers different portions of the biometrics (e.g., different areas of a finger) than other biometric data in the profile (e.g., template). The end result of this process is a gradual tuning of the matching process to the peculiarities exhibited by a user, thereby enhancing accuracy, speed and flexibility. This adaptation also accommodates the gradual changes in a subscriber's biometric data (e.g., fingerprints) over time.[0007]
• In another embodiment, the authentication process uses a challenge-response protocol. Using of the challenge-response protocol, neither the server nor the client transmit a full set of biometric data across the network during the authentication session. The server makes a copy of the user's biometric data and modifies the copy to generate a challenge template. The modifying can include eliminating some of the geometric data representing the biometric features (e.g., only including the x, y coordinates of a feature) and inserting fictitious data (e.g., random noise). The server transmits the challenge template to the client. The client receives the challenge template and compares the challenge template to a candidate set of biometric data. Based on the comparison, the client generates a response vector. The response vector can be, for example, a hash code. The client transmits the response vector back to the server. The response vector indicates the portions of the challenge template that did not match the candidate set of biometric data. The server, knowing what fictitious data was inserted into the challenge template can determine if the mismatches sufficiently match the fictitious data. If they do, the server can authenticate the user.[0008]
• In one aspect, the invention relates to a method for authentication using biometrics. The method comprises receiving a request for authentication of a user and receiving a first set of biometric data from the user. The method also comprises comparing the first set of biometric data with a second set of biometric data in storage; and modifying the second set of biometric data in storage based at least in part on the first set of biometric data, if the second set of biometric data sufficiently matches the first set of biometric data. In one embodiment, the method further includes replacing the second set of biometric data in storage with the first set of biometric data, if the second set of biometric data sufficiently matches the first set of biometric data.[0009]
• In another embodiment, the method further includes identifying one or more features in the first set of biometric data, matching at least a portion of the one or more features in the second set of biometric data and augmenting the second set of biometric data with the first set of biometric data based at least in part on the matched features. In another embodiment, the method further includes augmenting the second set of biometric data with features from the first set of biometric data not presently included in the second set of biometric data. In still another embodiment, the method further includes augmenting statistical data associated with the second set of biometric data based at least in part on the first set of biometric data. In yet another embodiment, the method further includes augmenting statistical by increasing the weighting of the matched features.[0010]
• In another embodiment, the method further includes normalizing and/or filtering the first set of biometric data. In another embodiment, the method further includes extracting from the first set of biometric data identifying features. The second set of biometric data may be stored in a supertemplate.[0011]
• In another aspect, the invention relates to a system for authentication using biometrics. The system preferably comprises a network interface, a storage module, an authentication module and an augmentor module. The network interface module may be configured to receive an authentication request requesting authentication of a user and to receive a first set of biometric data from the user. The storage module may be configured to store a second set of biometric data. In general, the authentication module is configured to compare the first set of biometric data with the second set of biometric data in storage. The augmentor module may modify the second set of biometric data in storage based at least in part on the first set of biometric data if the second set of biometric data sufficiently matches the first set of biometric data.[0012]
• In one embodiment, the augmentor module is further configured to replace the second set of biometric data in storage with the first set of biometric data if the second set of biometric data sufficiently matches the first set of biometric data. In another embodiment, the augmentor module is further configured i) to identify one or more features in the first set of biometric data, ii) to match at least a portion of the one or more features in the second set of biometric data and iii) to augment the second set of biometric data with the first set of biometric data based at least in part on the matched features. In still another embodiment, the augmentor module is further configured to augment the second set of biometric data with features from the first set of biometric data not presently included in the second set of biometric data.[0013]
• In another embodiment, the augmentor module is further configured to augment statistical data associated with the second set of biometric data based at least in part on the first set of biometric data. For example, the augmentor module may be configured to increase the weighting of the matched features. In another embodiment, the system further comprises a normalizer module configured to normalize the first set of biometric data. In still another embodiment, the system further comprises a filter module configured to filter the first set of biometric data. In yet another embodiment, the system further comprises an extractor module configured to extract identifying features from the first set of biometric data. The second set of biometric data may be stored in a supertemplate.[0014]
• In another aspect, the invention relates to an article of manufacture having computer-readable program portions embodied therein for authentication using biometrics. The article comprises computer-readable program portions for performing the method steps as described above.[0015]
BRIEF DESCRIPTION OF THE DRAWINGS
• The above and further advantages of the invention may be better understood by referring to the following description taken in conjunction with the accompanying drawing, in which:[0016]
• FIG. 1 is block diagrams of illustrative embodiments of a system to authenticate a user using augmented biometric data accordance with the invention;[0017]
• FIG. 2 is a block diagram of an illustrative embodiment of a supertemplate used to authenticate a user in accordance with the invention;[0018]
• FIG. 3 is a block diagram of another illustrative embodiment of a system to authenticate a user using augmented biometrics in accordance with the invention;[0019]
• FIG. 4 is a block diagram of another illustrative embodiment of a supertemplate used to authenticate a user in accordance with the invention; and[0020]
• FIG. 5 is a flow diagram of an illustrative embodiment of a process to authenticate a user using augmented biometrics in accordance with the invention.[0021]
DETAILED DESCRIPTION
• In broad overview, FIG. 1 illustrates an embodiment of a[0022]system100 to authenticate a user using augmented biometric data in accordance with the invention. Thesystem100 includes a first computing system (“a server node”)108 and a second computing system (“a client node”)112, all in communication with anetwork116. Theserver node108 and theclient node112 are in communication with the network usingcommunication channels120.
• For example, the[0023]network116 and thecommunication channels120 can be part of a local-area network (LAN), such as a company Intranet, a wide area network (WAN) such as the Internet or the World Wide Web or the like. Thenodes108 and112 communicate with thenetwork116 through thecommunication channels120 using any of a variety of connections including, for example, standard telephone lines, LAN or WAN links (e.g., T1, T3, 56 kb, X.25), broadband connections (ISDN, Frame Relay, ATM), wireless connections and the like. The connections can be established using a variety of communication protocols (e.g., HTTP(S), TCP/IP, SSL, IPX, SPX, NetBIOS, Ethernet, RS232, direct asynchronous connections, a proprietary protocol and the like). In one embodiment, theserver108 and theclient112 encrypt all communication when communicating with each other.
• The[0024]server node108 can be any computing device capable of providing the services requested by theclient node112. Particularly, this includes authenticating a user at theclient node112 using biometric data, as described in more detail below. Theserver node108 may include anetwork interface module124, anauthentication module128, anaugmentor module132 and astorage module135. The storage module135 (which may be, for example, persistent memory, one or more hard disks, optical drives and the like) can include atemplate136, in which a reference set of biometric data is stored. Theserver node108 can also include one or more optional modules that add additional features for the collection of biometric data and are used inpath138, i.e., between thenetwork interface module124 and theauthentication module128. For example, theserver108 can include anormalizer module142, afilter module144 and/or anextractor module146. The modules discussed throughout the specification are implemented as a software program and/or a hardware device (e.g., ASIC, FPGA, processor, memory, storage and the like). In one embodiment, one or more of theoptional modules142,144 and/or146 may be included on theclient112 instead of or in addition to theserver108. Placing the one or more of theoptional modules142,144 and/or146 on theclient112 distributes the processing task and lowers needed bandwidth on thenetwork116.
• For clarity, FIG. 1 depicts[0025]server node108 as a single server. It is to be understood, however, that theserver node108 can also be implemented, for example, distributed on portions of several (i.e., more than two) servers. Theclient node112 can be any computing device (e.g., a personal computer, set top box, wireless mobile phone, handheld device, personal digital assistant, kiosk, etc) used to provide a user interface to access theserver108. Theclient node112 receives biometric data from a biometric input device160 (e.g., a fingerprint scanner, a retina scanner, a thermal imager, a skin spectrometer, a voice print analyzer, a digital camera and the like).
• To use the[0026]system100, a user170, also referred to as a subscriber, registers that user's170 biometric data with thesystem100. In the illustrated embodiment, theclient112 receives biometric data from thebiometric input device160. The biometric data can include, for example, data associated with the individual's fingerprint(s), facial characteristics, voice and the like. Thesystem100 stores a set of biometric data associated with the user170 in thestorage module135. In one embodiment, the biometric data is stored using an alias (e.g., a unique identifier with no personal or other type of information that can identify an individual), so that if the security of thestorage module135 is compromised, the biometric data cannot be associated with a particular individual.
• In general overview, when the user[0027]170 requests a service over thenetwork116 that requires authentication, theclient device112 receives a candidate set of biometric data from thebiometric input device160 and transmits it to theserver node108. Thenetwork interface module124 receives the candidate set of biometric data and transmits it to theauthentication module128. Theauthentication module128 retrieves a reference set of biometric data associated with the user170 from thestorage module135. If the candidate set of biometric data sufficiently matches the reference set of biometric data, theauthentication module128 authenticates the user as the registered individual.
• To authenticate, the[0028]authentication module128 and/or theoptional modules142,144 and/or146 process the received candidate set of biometric data to extract the unique features that distinguish one set of biometric data (e.g., fingerprint) from another. For example, thenormalizer module142 normalizes the biometric data into a format used by theauthentication module128 and stored in thestorage module135. The normalization can include, for example, a translation algorithm, a transformation algorithm and the like. The normalization allows the biometrics data to be converted into a standard image suitable for subsequent processing and preferably includes geometric processing to adjust for size differences between sensors, orientation adjustments to invert or rotate images, density adjustments to correct for number of gray levels/dynamic range and sampling adjustments to account for different sensor resolutions. This allows theclient device112 to interface with different types of biometric input devices160 (e.g., fingerprint readers produced by different manufacturers and having diverse capture resolutions or characteristics) without the need to re-register the user170 or change the format of the biometric data in thestorage module135.
• The[0029]filter module144 filters the received candidate set of biometric data. The filtering can include standard filtering algorithms for correcting blurring of the image, for removing random noise in the image and the like. For example, all captured scans can be checked for partial or blurred prints that exhibit greater than expected amount of change between consecutive frames as well as contrast. Images that exhibit excessive blur can be rejected. Contrast issues can be resolved by asking the user to press down to make better contact with the sensor. Image processing software may be used to enhance the quality of the image and involve signal averaging, noise filtering, ridge/valley enhancement as well as gray scale equalization. The filtering can also include filtering algorithms needed because of the type of thebiometric device160 or the type of user features thebiometric device160 uses. The filtering can also include filtering algorithms based on the type of image (e.g., grainy, wet, fine grain and the like), the finger type and/or personal biometric characteristics (e.g., sex, age and the like). In an embodiment where thefilter module144 is implemented on theclient112, the filter module114 operates in conjunction with thebiometric input device116 to perform blur removal, finger detection and time based enhancements. For example, two or more scans may be taken to ensure the user170 has placed a stable finger (not moving) on the sensor. A difference is then taken between subsequent scans to ensure consistency between the two scans. With noisy sensors, thefilter module144 may integrate consecutive images to reduce the noise level in the captured image.
• The[0030]extractor module146 extracts the geometric data representing biometric features and/or minutiae from the candidate set of biometric data. In an embodiment where theextractor module146 is implemented on theclient112, theextractor module146 transmits the results to theauthentication module128 using thenetwork116. Biometric data, for example in the case of fingerprints, can be divided into global features that are spatial in nature and local features that represent details captured in specific locations. The geometric data can include, for example in the case of fingerprints, the locations (e.g., x, y coordinates) of the features, the type of feature (e.g., ridge ending, bifurcation and the like), the angular data of the features, the slope of the ridge, the neighborhood ridge counts and/or the like. Once the geometric data is processed, theauthentication module128 compares the data of the reference set of biometric data stored in thestorage module135 with the candidate set of biometric data to produce a goodness of fit or confidence of match by examining the local features on a minutia by minutia basis. To calculate the goodness of fit, theauthentication module128 determines the best spatial alignment between the location of minutiae points within the reference set of biometric data and corresponding minutiae points within the candidate set of biometric data. Determining the best spatial alignment involves, for example, finding the rotation angle that produces the greatest number of matching points. Matching can be a relative term, meaning the points are close to each other within some predefined distance. The determining process preferably accommodates both spatial and rotational displacement between the reference set of biometric data and the candidate set of biometric data. This may be accomplished, for example, using a spatial correlation algorithm in which the features of the candidate set of biometric data are translated and rotated about a test alignment point and then compared against the features in the reference set. Different alignment points and rotation angles are tested to determine the lowest difference between the candidate and reference feature set. Once the differences between the local features at each of the matching minutiae points are minimized, theauthentication module128 sums the goodness of fit.
• The[0031]authentication module128 determines the sufficiency of the match by statistically analyzing the goodness of fit for local features at each of the matching minutiae points and determining whether the probability that they come from the same individual is above a certain predetermined threshold. In one embodiment, an administrator of thesystem100 sets the predetermined threshold. The predetermined threshold determines both the false acceptance rate (i.e., the probability that theauthentication module128 will incorrectly authenticate a user) and the false rejection rate (i.e., the probability that theauthentication module128 will incorrectly reject the user when that user is in fact the registered individual). The administrator sets the predetermined threshold such that the false acceptance rate and the false rejection rate are both acceptable to the users of thesystem100.
• In addition, with the sufficient match, the[0032]authentication module128 transmits the candidate set of the biometric data to theaugmentor module132, which in turn modifies the current reference set of biometric data (e.g., template136) using the candidate set of biometric data. The modification can include several different aspects of the reference biometric data. For example, one aspect is the spatial aspect (e.g., the associated data representing geometric features) of the reference set of the biometric data. Another aspect is the statistical aspect (e.g., the weighting and/or confidence level of features) of the reference set of the biometric data.
• FIG. 2 illustrates an exemplary embodiment of a[0033]supertemplate200 used to authenticate a user in accordance with the invention. Thesupertemplate200 represents a set of biometric data corresponding to a complete set of biometric data. For example, in a fingerprint system, thesupertemplate200 represents the complete set of biometric data for one digit. Superimposed on thesupertemplate200 are a first set ofbiometric data205, a second set ofbiometric data210, a third setbiometric data215 and a fourth set of biometric data220. As illustrated, the sets ofbiometric data205,210,215 and220 are smaller in size than thesupertemplate200. In one embodiment, the sets ofbiometric data205,210,215 and220 representtemplates136. Thesupertemplate200 can comprise one ormore templates136. The size of the sets ofbiometric data205,210,215 and220 are based on thebiometric input device160. For example, the size of the scanner, the size of the local memory and the like. It is noteworthy that even if the scanner is large enough to cover the entire finger thesupertemplate200 can accumulate additional information frommultiple templates205,210,215 and220 to generate more accurate statistics for the features.
• For an illustrative example of the modifying process, a reference set of biometric data is the[0034]supertemplate200 and a candidate set of biometric data for a first authentication request is equivalent to the biometric data represented in the first set ofbiometric data205. As described above, upon a sufficient match, theaugmentor module132 modifies thesupertemplate200 using the candidate set of biometric data. For the sufficient match, theauthentication module128 matches features of the candidate set of biometric data with features of the reference data included in thesupertemplate200. Theaugmentor module132 aligns those matched features to determine how the candidate set of biometric data fits into thesupertemplate200.
• When the[0035]augmentor module132 determines the alignment, theaugmentor132 modifies thetemplate200 using the candidate set of biometric data. The results are that in this illustrative example, the area indicated as the first set ofbiometric data205 is modified with the candidate set of biometric data. In one embodiment, theaugmentor module132 modifies by replacing the features in the existing reference biometric data in the area indicated as the first set ofbiometric data205 with the candidate set of biometric data received by theclient112. In this way, the system accommodates feature changes that occur over time (e.g., due to aging of the user). In another embodiment, theaugmentor module132 augments the existing reference biometric data by adding in those features of the candidate set of biometric data that are not matched and/or not presently included in the reference set of biometric data. This allows thesystem100 to build a fuller biometric representation than would be possible with, for example, a single scan by thebiometric input device160. Augmented in this fashion, thesupertemplate200 can evaluate scans covering different portions of, for example, the user's fingerprint, increasing the system's tolerance for variation without sacrificing accuracy (i.e., the number of feature points matched).
• Continuing with the illustrative example of the modifying process, a candidate set of biometric data for a second authentication request is equivalent to the biometric data represented in the second set of[0036]biometric data210. As described above, upon a sufficient match, theaugmentor module132 aligns those matched features to determine how the candidate set of biometric data fits into thesupertemplate200. Once theaugmentor module132 determines the alignment, theaugmentor132 modifies thetemplate200 using the candidate set of biometric data. The results are that in this illustrative example, the area indicated as the second set ofbiometric data210 is modified with the candidate set of biometric data. Similarly in subsequent authentication requests, theaugmentor module132 modifies the areas indicated as the third and fourth sets of biometric data,220,225 respectively.
• As described above, in addition to the modification of features, the[0037]augmentor module132 also modifies the statistical parameters of the reference set of biometric data. Each time features in a candidate set of biometric data match the features of the reference set biometric data, theaugmentor module132 increases the weighting and/or confidence level of those matched features. Thearea230, indicated by shading, represents the overlap of all of the sets of biometric data (205,210,215 and225). The weighting and/or confidence level of the matched features in thisarea230 is the highest, as it has been reinforced by the redundant presence of the matched features in each of the four received candidate sets of biometric data. The closeness of the match can also affect the value of the weighting and/or confidence level. For example those features that directly overlap with two candidate sets of biometric data have a higher weighting and/or confidence level than those features that are close, but have some small distance between them.
• In broad overview, FIG. 3 illustrates another embodiment of[0038]system100′ to authenticate a user using augmented biometric data in accordance with the invention. Theserver node108′ of thesystem100′ includes anetwork interface module124′, anauthentication module128′, astorage module135, having atemplate136′ stored therein, and amodification module320. Theclient node112′ of the system includes aclient comparator module330.
• To use the[0039]system100′, the user170 registers that user's170 biometric data with thesystem100′, as described above. For authenticating, theserver108′ andclient112′ use a challenge-response protocol that does not transmit a full set of biometric data across thenetwork116. This challenge-response protocol modifies a portion of the set of biometric data sent across thenetwork116 so that if intercepted by someone, it is not usable in its modified state. FIG. 4 depicts asupertemplate200′ that thesystem100′ employs to implement the challenge-response protocol. Thesupertemplate200′ includes achallenge template405 that represents a set of biometric data. In one embodiment, thechallenge template405 is equivalent in area to thetemplate136′. As described above, thetemplate136′ varies in size and is at least a portion of the supertemplate200′. Thechallenge template405 includes afirst portion410 and asecond portion420. As illustrated, thefirst portion410 and thesecond portion420 include random feature data, as described in more detail below.
• FIG. 5 illustrates an embodiment of a[0040]process500 to authenticate a user170 using the challenge-response protocol, asystem100′ as depicted, for example, in FIG. 3 and achallenge template405 as depicted, for example, in FIG. 4. In operation, theclient112′, in response to a user170 action, generates (step505) a request. The request can be an authentication request directly from theclient112 to authenticate the user170. The request can also be a service request for a certain service (e.g., execution of an application program, access to a financial or medical database, access to an electronic vault with which the user170 is associated, download of data and/or an application program, and the like) provided by a server on the network (e.g.,108′ or a different application server). In that case, the server providing the requested service transmits a request for authentication to theauthentication module128′.
• In response to the authentication request, the[0041]modification module320 copies (step510) thetemplate136′ of the reference biometric data associated with the user170. Themodification module320 generates (step515) modification data and uses this modification data to modify (step520) the copy of the template to generate achallenge template405. For an illustrative example, themodification module320 copies (step510) at least a portion of the geometric data contained within thereference template136′, for example, the x, y coordinates of the features. To generate the modification data, themodification module320 generates random x, y locations and thereby generates random modified features at these locations in thechallenge template405. In another embodiment, the modification data is not random but generated by an algorithm that is dependent on the biometric data, thus creating different modification data for different users. Themodification module320 modifies (step520) the copy of the reference template (i.e., challenge template405) by inserting the modification data into thechallenge template405, for example at the random x, y locations. Themodification module320 can also create the modification data used for thechallenge template405 by combining features from other users or other fingers to create a composite that is similar to real data because the modification data is based on real biometric data. For example, themodification module320 can create the composite modification data from other users and then align the end points when inserting this composite modification data inportion410, so it looks like real data, but would not be matchable without knowing which areas were false.
• For clarity and illustration only, the modification data for this particular request of the illustrated[0042]process500 fall within the first and second portions,410 and420 respectively, of thechallenge template405. Of course, if the modification data were random, then the modified x, y coordinates would be distributed randomly throughout the challenge template area. In another embodiment, themodification module320 can insert random noise in portions of thechallenge template405, for example, in the first and second portions,410 and420 respectively. Once themodification module320 generates thechallenge template405, theserver108′ transmits thechallenge template405 to theclient112. As stated above, with random data inserted in thefirst location410 and thesecond location420, even if thechallenge template405 is copied by an eavesdropper, thechallenge template405 is not usable because the biometric data in those locations will not match reference biometric data (e.g.,reference template136′) stored for that user in biometric authentication systems.
• The[0043]comparator module330 of theclient112′ receives (step530) a set of candidate biometric data from thebiometric input device160. Thecomparator module330 compares (step535) the candidate set of biometric data with the receivedchallenge template405. Thecomparator module330, for example, can spatially align the candidate set of biometric data with thechallenge template405, maximizing the number of matching features, and then calculate a degree of overlapping (i.e., matching) of the features at various x, y coordinates. Thecomparator module330 generates (step540) a response vector, for example, listing the x, y coordinates and the degree of matching. Another format can include the actual candidate features found in all matching areas. Other formats for the response vector include listing the x, y coordinates that are above (or below) a certain threshold, listing the x, y coordinates with no matching features, generating a hash code using thechallenge template405 and the candidate set of biometric data, and the like. Theclient112′ transmits the response vector back to theserver108′. The transmitted response vector does not include a full set of biometric data, so it is not usable if someone intercepts it.
• The[0044]authentication module128′ receives the response vector and compares (step550) the response vector with the modification data. Theauthentication module128′ determines (step555) if the comparison between the response vector and the modification data indicates that there is a sufficient match, or in other words, that the user170 is, to a statistical degree of certainty, the registered individual. If theauthentication module128′ determines that the comparison indicates there is not a sufficient match, theauthentication module128′ denies (step560) the user1780 as the registered individual. If theauthentication module128′ determines that the comparison indicates there is a sufficient match, theauthentication module128′ authenticates (step565) the user1780 as the registered individual.
• Ideally, when the user[0045]170 is the registered individual, the mismatches identified in the response vector coordinate with the modification data in thefirst location410 and thesecond location420, while the features outside of these locations match to a high degree of probability. Deviations from this ideal can be caused by noise introduced by thebiometric input device160, different sizes of the candidate set of biometric data and thechallenge template405, rotation and/or motion of the user's finger while scanning, and the like. As described above, theauthentication module128′ statistically analyzes the mismatches, accounting for those due to the modification data, and determines to a statistical certainty whether the matches indicate that the user170 is the registered individual.
• Equivalents[0046]
• The invention can be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The foregoing embodiments are therefore to be considered in all respects illustrative rather than limiting on the invention described herein. Scope of the invention is thus indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.[0047]

Claims (22)

What is claimed is:

1. A method for authentication using biometrics, the method comprising:

receiving an authentication request requesting authentication of a user;

receiving a first set of biometric data from the user;

comparing the first set of biometric data with a second set of biometric data in storage; and

if the second set of biometric data sufficiently matches the first set of biometric data, modifying the second set of biometric data in storage based at least in part on the first set of biometric data.

2. The method ofclaim 1 wherein the modifying step comprises, if the second set of biometric data sufficiently matches the first set of biometric data, replacing the second set of biometric data in storage with the first set of biometric data.

3. The method ofclaim 1 wherein the modifying step further comprises:

identifying one or more features in the first set of biometric data;

matching at least a portion of the one or more features in the second set of biometric data; and

augmenting the second set of biometric data with the first set of biometric data based at least in part on the matched features.

4. The method ofclaim 3 wherein the augmenting step further comprises augmenting the second set of biometric data with features from the first set of biometric data not presently included in the second set of biometric data.

5. The method ofclaim 3 wherein the augmenting step further comprises augmenting statistical data associated with the second set of biometric data based at least in part on the first set of biometric data.

6. The method ofclaim 5 wherein the step of augmenting statistical data further comprises increasing the weighting of the matched features.

7. The method ofclaim 1 further comprising normalizing the first set of biometric data.

8. The method ofclaim 1 further comprising filtering the first set of biometric data.

9. The method ofclaim 1 further comprising extracting from the first set of biometric data identifying features.

10. The method ofclaim 1 wherein the second set of biometric data is stored in a supertemplate.

11. A system for authentication using biometrics, the system comprising:

a network interface module configured to receive an authentication request requesting authentication of a user and to receive a first set of biometric data from the user;

a storage module configured to store a second set of biometric data;

an authentication module configured to compare the first set of biometric data with the second set of biometric data in storage; and

an augmentor module configured to modify the second set of biometric data in storage based at least in part on the first set of biometric data if the second set of biometric data sufficiently matches the first set of biometric data.

12. The system ofclaim 11 wherein the augmentor module is further configured to replace the second set of biometric data in storage with the first set of biometric data if the second set of biometric data sufficiently matches the first set of biometric data.

13. The system ofclaim 11 wherein the augmentor module is further configured i) to identify one or more features in the first set of biometric data, ii) to match at least a portion of the one or more features in the second set of biometric data and iii) to augment the second set of biometric data with the first set of biometric data based at least in part on the matched features.

14. The system ofclaim 13 wherein the augmentor module is further configured to augment the second set of biometric data with features from the first set of biometric data not presently included in the second set of biometric data.

15. The system ofclaim 13 wherein the augmentor module is further configured to augment statistical data associated with the second set of biometric data based at least in part on the first set of biometric data.

16. The system ofclaim 15 wherein the augmentor module is further configured to increase the weighting of the matched features.

17. The system ofclaim 11 further comprising a normalizer module configured to normalize the first set of biometric data.

18. The system of claim111 further comprising a filter module configured to filter the first set of biometric data.

19. The system ofclaim 11 further comprising an extractor module configured to extract from the first set of biometric data identifying features.

20. The system ofclaim 11 wherein the second set of biometric data is stored in a supertemplate.

21. An article of manufacture having computer-readable program portions embodied therein for authentication using biometrics, the article comprising:

a computer-readable program portion for receiving an authentication request requesting authentication of a user;

a computer-readable program portion for receiving a first set of biometric data from the user;

a computer-readable program portion for comparing the first set of biometric data with a second set of biometric data in storage; and

a computer-readable program portion for modifying the second set of biometric data in storage based at least in part on the first set of biometric data if the second set of biometric data sufficiently matches the first set of biometric data.

22. The article ofclaim 21 wherein the modifying step further comprises:

a computer-readable program portion for identifying one or more features in the first set of biometric data;

a computer-readable program portion for matching at least a portion of the one or more features in the second set of biometric data; and

a computer-readable program portion for augmenting the second set of biometric data with the first set of biometric data based at least in part on the matched features.

Images