Image analysis – Applications – Personnel identification
Reexamination Certificate
1997-11-17
2001-12-11
Bella, Matthew C. (Department: 2621)
Image analysis
Applications
Personnel identification
C382S124000, C382S126000, C382S275000, C382S318000, C348S218100, C348S229100, C358S446000, C358S463000, C358S483000, C358S504000
Reexamination Certificate
active
06330345
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to sensor devices, and more particularly to a method and apparatus that accounts for device variabilities and adjusts for variable conditions that are present when imaging an individual's biometric feature and in particular, a fingerprint image.
BACKGROUND OF THE INVENTION
Biometric-oriented personal identification techniques are becoming increasingly important in protecting personal property, such as laptop computers and cellular phones, preventing credit card and calling card fraud, limiting access to security areas, computers and information, and ensuring security for electronic commerce. Biometric identification techniques employ physical traits, measurements and other characteristics specific to an individual. These characteristics include, but are not limited to, voice prints, hand prints, fingerprints, retina patterns, and signature analysis. Typically, biometric identification and verification techniques compare an individual's stored biometric data against newly obtained biometric data when the individual desires use of a protected item, or access to a protected area or information. Due to its inherent nature, biometric data has the advantage of always being available for user identification and verification. However, also by its inherent nature, due to the vagaries of the human body, biometric data is often difficult to acquire in a consistent manner to yield an unambiguous measure of identity.
The fingerprint biometric is probably the most widely used and researched biometric identification technique. Existing technology allows the relevant features of a fingerprint to be represented in a few hundred bytes of data. Furthermore, the computer hardware required for recording and comparing fingerprint data can be centralized and accessed through a telecommunications network thereby allowing costs to be amortized across many transactions.
The disadvantage of biometric identification and verification, and in particular fingerprint identification, is in acquiring an accurate image of the fingerprint each time an individual desires to use or access the protected item. The problems associated with acquiring an accurate fingerprint image include sensor device dependent variables, such as hardware defects and deterioration, and individual conditions, such as the moisture content and temperature of the feature being imaged.
Accordingly, there is needed a sensor device employing methodology that accounts for sensor device variability, human biometric variability and adjusts for variable conditions that are present when imaging an individual's biometric features.
SUMMARY OF THE INVENTION
The present invention is an apparatus and a method that increases accuracy and reliability of biometric identification and verification techniques. An increase in accuracy and reliability is achieved by automatically accounting for sensor device variations and adjusting for variable conditions that are present when preparing to capture an image of an individual's biometric feature. The present invention optimally adjusts for the condition of an individual's biometric features and captures multiple images at different optimized sensitivity settings in a relatively small time frame, such that capture and adjustment processing is transparent to the individual.
In an exemplary embodiment, the present invention automatically determines a set of default settings and diagnostic data for a sensor device first based on a blank image. At least one set of optimized settings is then determined by automatically adjusting the set of default settings based on the condition of the individual's biometric feature. Images are then captured of the feature, where the number of images taken corresponds to the number of optimized sets. The methodology corrects the captured image or images with the previously stored diagnostic data. If multiple images are captured, the present invention combines the corrected images to form a resultant image that is of higher quality than any of the single images.
Advantageously, the invention automatically determines default settings that provide for a substantially uniform and high contrast image by accounting for device variations. The default settings are adjusted for the variable conditions of the feature being imaged, for example, whether a finger is too wet or too dry, by examining a specific set of regions of the image data rather than the entire image.
In one exemplary embodiment of the present invention, multiple images are captured at different sensitivity levels to permit greater accuracy in identification. The captured images are corrected with previously determined and stored diagnostic data, such as non-functional pixel regions. The multiple images are then combined by using a weighted and intensity-translated combination of regions from the different images to form a resultant image.
REFERENCES:
patent: 3949162 (1976-04-01), Malueg
patent: 4194220 (1980-03-01), Frame
patent: 4343021 (1982-08-01), Frame
patent: 4408231 (1983-10-01), Bushaw et al.
patent: 4525741 (1985-06-01), Chahal et al.
patent: 4638365 (1987-01-01), Kato
patent: 4639781 (1987-01-01), Rucci et al.
patent: 4933976 (1990-06-01), Fishbine et al.
patent: 4991127 (1991-02-01), Crookshanks
patent: 5047861 (1991-09-01), Houchin et al.
patent: 5151945 (1992-09-01), Lee et al.
patent: 5325442 (1994-06-01), Knapp
patent: 5467403 (1995-11-01), Fishbine et al.
patent: 5524069 (1996-06-01), Inoue
patent: 5544338 (1996-08-01), Forslund
patent: 5613013 (1997-03-01), Schuette
patent: 5692065 (1997-11-01), Prakash et al.
patent: 5838306 (1998-11-01), O'Connor et al.
patent: 5914748 (1999-06-01), Parulski et al.
patent: 5920640 (1999-07-01), Salatino et al.
patent: 5942761 (1999-08-01), Tuli
patent: 6177958 (2001-01-01), Anderson
patent: 0813164 (1997-12-01), None
O'Gorman Lawrence
Russo Anthony P.
Bella Matthew C.
Chawan Sheela
Gibbons Del Deo Dolan Griffinger & Vecchione
Veridicom, Inc.
LandOfFree
Automatic adjustment processing for sensor devices does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Automatic adjustment processing for sensor devices, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Automatic adjustment processing for sensor devices will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2556489