CA2083380C - Method for identifying individuals from analysis of elemental shapes derived from biosensor data - Google Patents
Method for identifying individuals from analysis of elemental shapes derived from biosensor dataInfo
- Publication number
- CA2083380C CA2083380C CA002083380A CA2083380A CA2083380C CA 2083380 C CA2083380 C CA 2083380C CA 002083380 A CA002083380 A CA 002083380A CA 2083380 A CA2083380 A CA 2083380A CA 2083380 C CA2083380 C CA 2083380C
- Authority
- CA
- Canada
- Prior art keywords
- individual
- image
- thermal image
- thermal
- elemental
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/01—Measuring temperature of body parts ; Diagnostic temperature sensing, e.g. for malignant or inflamed tissue
- A61B5/015—By temperature mapping of body part
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/117—Identification of persons
- A61B5/1171—Identification of persons based on the shapes or appearances of their bodies or parts thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C9/00—Individual registration on entry or exit
- G07C9/20—Individual registration on entry or exit involving the use of a pass
- G07C9/22—Individual registration on entry or exit involving the use of a pass in combination with an identity check of the pass holder
- G07C9/25—Individual registration on entry or exit involving the use of a pass in combination with an identity check of the pass holder using biometric data, e.g. fingerprints, iris scans or voice recognition
- G07C9/253—Individual registration on entry or exit involving the use of a pass in combination with an identity check of the pass holder using biometric data, e.g. fingerprints, iris scans or voice recognition visually
Abstract
A method and apparatus for identifying individuals is characterized by the analysis of elemental shapes derived from biosensor data presented in n dimensions, particularly facial thermograms. A thermal image of the individual's face is generated (4) in accordance with the thermal energy patterns thereof. The thermal image is converted to a digital representation (8) comprising a matrix of pixels, the intensity of each pixel corresponding with the level of thermal energy of a corresponding portions of the image. The digital representation of the image is normalized for size and equalized to a standard histogram to produce an image containing elemental shapes comprising contours corresponding with unique structural features of the individual (12).
Description
METHOD FOR IDENTIFYING INDIVIDUALS FROM ANALYSIS
OF ELEMENTAL SHAPES DERIVED FROM BIOSENSOR DATA
BACKGROUND OF THE INVENTION
The problem of ascertaining the identity of an individual or verifying whether an individual is the person he claims to be is a common one continually fa.ed by individuals, businesses, and governments Me'; od for positive identification of an individual usually include reliance upon knowledge of restricted information (such as a password), possession of a restricted article (such as a passkey), or physical appearance (such as matching a reference photo).
Security based upon knowledge or possession may be compromised without discovery since r_he information or article may be extorted from its rightful owner. The third methodology, commonly referred to as biometric techniques, is considered less vulnerable to m~:~take.n identity. The best known biometric technique .~md tr~F:: one considered most secure is fingerprint analysis. Hou~Pver, other biometric techniques exist. The present inventyor~ relates to an identification technique utilizing thermal imaging.
Skin temperature is a result of the balance between the heat flow from tissue to the skin surface and the heat flow from the surface to the surroundings. The basic mechanisms for transport of heat from the tissues to the skin surface include conduction throughout the tissues and convection by the flow of blood from the warm art:~rial supply through the capillary bed into the venous plex~~s. Metabolic heat generation is a second order influence compared to the convection transport mechanism of the blood perfu.,ion. Thus in the absence of active superficial veins, the diffuse avascular skin temperature is directly related to the blood perfusion per unit volume of tissue, including the skin itself as well as tissues approximately 2 to 3 em deep to the skin. Active superficial veins are also a possible indicator of regional blood flow. These azn dr..~~mstrat~d as well-defined patterns of increased skin temperature. Skin temperature immediately overlying the vein is only a few tenths of a degree Celsius lower than the temperature of the blood flowing in the vein itself .
A variation of 3.5 degrees Fahrenheit is usual across a human facial surface. In the ::ase of severe medical abnormalities, or when the ambient temperature is severe, the range may be twice as large, and its average value may vary.
The heat energy given off by the face is changed to electrical energy and greatly amplified by the electronic circuitry of a thermal imaging system. Following amplification, the electrical energy is digitized. A facial "thermogram" is the resultant two-dimensional image obtained by scanning each section of the face onto one or more detectors which are sensitive in the infrared range. The most commonly used medical thermal imagers produce an image which comprises approximately 60,000 individual heat recordings. Most commonly used scanning infrared cameras do not accurately measure vein temperature because the modulation transfer function of the optics of these cameras produces a measurement that is related to the average temperature over almost 1 cm2 of scanning surface. Thus the size of the vein will influence the indicated temperature.
Because a thermograph operates at a distance from the subject and detects and records only radiant heat spontaneously emitted from the body surface, it constitutes a painless, non-invasive, passive method of recording body surface temperatures. In general, the nose, cheeks, ears, and the avascular corneas of the ey:s are cool, as are the hair-insulated eyebrow, eyelash, and scalp regions. The anterior portion of the forehead is warm and has a narrow range of temperature variation over its surface. Heat is normally trapped by skin folds, at the canthi and along the free margins of the eyelids. EssentiGlly symmetrical facial thermograms are found almost without exception in healthy subjects.
In accordance with the ~.~resent invention, facial thermograms are processed to extract elemental shapes therefrom. the elemental shapes -- like a fingerprint --are unique to an individual. Thus, the identity of an individual may be asce---tamed by comparing current elemental shapes with those from a z~~erence.
BRIEF DESCRIPTION OF THE PRIOR ART
Individuals identify one ar_other primarily by recognizing their faces. It is therefore understandable that systems would be developed which would attempt to replicate how persons recognize one another. Various patents have been issued which utilize known photographic or other visual images for comparison with unknown imagea in order to automate the recognition task. In the facial curve method of the Rothfjell U.S. Patent No. 3,805,238, facial curves (two or three dimensional) of an individual t~~ken from fullface, forty-five degree angle, and profile (ninety degree angle) are stored on an identity card or in a data processing machine.
Identification of the individual is performed by first visually comparing the photographs on the identity card to the individual presenting the card and then taking an image of the individual and automatically comparing by computer the curves stored on the identify card with curves derived from the image. In order to properly compare such an image with the computer data, however, the individual must be positioned and aligned in front of a glass screen.
It is also known in the art to measure facial features utilizing ratios obtained from a set of measurements taken from a facial image as disclosed in the Tal U.S. patent No.
4,975,969. The use of binary images obtained by thresholding facial images collected under relatively uncontrolled conditions and finding the best :r,atch among small groups of known indi~=iduals is disclosed i_~ the U.S. patent to Lu No.
4,858,000.
The time ~equ~ -~ed fear enrol-i mer_'~ and verification, the difficulty in automat-i.r:g the tec}rniques, and variabilities resulting from lighting conditions and shadows are major Y:~;? ~.
_ ~Y
limitations to the ro~.aine, unattended use of these and other facial recognition systems whicY_~~tilize photographs or video images. In addition, sucr~ systems are vulnerable to forgery through disguise and use ~f photographs or mannequins.
Identification tecr~niques utilizing infrared thermal imaging are known in the art as evidenced by the U.S. patent to Rice No. 4,699,149. This pate~~c is predicated upon the discovery that the arrangement _- veins in a local portion of an individual's body is a uniiue characteristic of that individual. IdentificatioYi is pe-formed by linearly scanning a region of skin at the wris, and detecting the position of subcutaneous blood vessels by measuring a parameter such as the reflection of incident radiation. A transducer is constrained to follow a particular path, and the resulting linear scan measurements arE thresholded to produce a binary image which constitutes the identifs:ation signature. The measured signature and predetexo,inec~ signature at the same location are compared.
A number of drawbacks are inherent in the Rice system.
for example, active illumination cY heating of the area being scanned is required, as is identical replication of the small skin area selected for scanning. These requirements may cause discomfort to the individual. Moreover, identification can not be verified from a distance as is necessary for security access systems.
The present invention was de~reloped in order to overcome these and other drawbacks of the prior identification techniques by providing a passi~-~ identification method wherein unique elemental shapes are extracted from a thermal image of the individual.
SUMMARY OF THE INVENTION
Accordingly, it is a prirrrary object of the present invention to provide a method for identifying individuals from biosensor data. A thermal image of a portion of the individual's body is generateu in accordance with the thermal energy patterns thereof. T:~e therm~:_ image is processed to ;: ~ ~'~ ° .., produce a digital representation thereof. The digital representation includes a ma rix of pixels and the intensity of each pixel correspor_~~s w~' ~ ~ tr-:.e lewal of thermal energy of a corresponding portior:.~f ne image. The digital image is normalized in order to elim_~.~_ate variables and produce a normalized image containing elemental shapes comprising contours corresponding with unique structural features of the individual. Normalization is accomplished by cropping the thermal image to a cropped area of maximum uniqueness of an individual relative to others, sucr~ as the eye and nose area of the individual°s face, stand~.rd~zing the image for size, and equalizing the image to a standard histogram. The processed image is stored for subsequent identification of the individual.
According to another object or the invention, a region of the image containing specific elemental shapes of interest is defined by geometry, mathematics, or as a bit-map replica of the region in order to provide ~_ signature of the individual.
Correlation techniques are used to compare a previously defined reference signature vi an individual with a currently defined signature to cor:fir.; the individual's identify.
BRIEF DESCRIPTION OF THE FIGJRES
Other objects and advantages of the invention will become apparent from a st~~dy of the fol~_r~wing specification when viewed in the light of the ac~~ompar_ying drawing, in which:
Fig. 1 is a block diagram of the apparatus used to create an identification signature of ar! individual;
Fig. 2 is a flow chart illustrating the steps for creating the identification signature of an individual from biosensor data in accordance with the invention;
Fig. 3 is a flow chart illustrating the steps for identifying an individual to_ascertain whether entry of the individual to a secure area is to be provided;
Fig. 4 is a digitized thermal image of an individual's face;
Fig. 5 is the image of Fig. 4 including axes used for ' r . :v,~. z s ..
-G-cropping;
Fig. 6 is a digitized th:~rmal contcur image of the image of Fig. 4 after normalizatic~z for size;
Fig. 7 is the image of F~e~. 5 following histogram equalization;
Fig. 8 illustrates correlat~om of two images to identify a match of certain features thereof;
Fig. 9 is a detailed illast._atir~n of do elemental shape;
and Figs. 10 and 11 ill~_zst~ate the correlation results of good and bad match areas, respectively.
DETAILED DESCRIPTION
The invention relates to ~. method and apparatus for uniquely identifying individu_.~s through analysis of elemental shapes derived from biosensor data p-cesented in n dimensions.
In order to conform the identity of an individual, a reference identification for that individual must be obtained.
This is performed by enrol~.ing or storing a thermogram image of the individual in a storage ~?ev~ce such as a storage card or a memory for subsequent comparison. In this manner, when the individual presents himself for identification, such as when he desires to gain entry to a secure area, the current thermogram image of the in,~ivi~.ual can be compared with the stored reference to confirm the individual's identity. the apparatus for generatir_g a thermogram image of the individual will be described with reference to Fig. 1.
A position detector 2 is ~~se.to determine whether the individual is properly framed ir~ front of an image sensor 4.
In the preferred embodiment, the face of the individual is that portion positioned in front of the sensor from which the thermogram is generated. As used herein the term face includes all portions c~f t.ne head imaged by the sensor 4. The degrees of tilt, twist, and tip can be made with reference to polar coordinates establishing ~'ie ''faceball" as will be described below. The preferred emboc~imemt results in a standard 30° upwards angle t~ th= fa~:e during imaging, which reduces the possibility of impinging hair or clothing on the minimum portion of the f~:ce r~eauired for most accurate identification, that p~~ri io:_ ~..eing the swath occurring between the approximately noriz~mcal 1-nes established by the uppermost points of the eyebrows and by the nostrils or lowest portion of the nose.
The image sensor 4 preferably comprises a platinum silicide staring array camera although a ~-ariety of other infrared imagers of e-they the array or scanning type may be used. Furthermore, other t:.~pes ~n n-dimensional biometric sensors may be used including for example magnetic resonance imagers, digital angiographs, computed tomographs, sonographs, radioisotope imagers, and X-_~ay devices. Infrared imagers producing emissions in the 3-6 or 8-14 micron ranges are preferred since thermal data from the individual may be collected passively and instantaneously with no risk to the individual.
The image sensor 4 produces an analog image of the individual's face wherein the intensity level is direction proportional to the absolute temperature at points across the face, where the size of the points depends on the imager used.
The analog image is sent to ar_ image processor 6 for analysis. The processor inc~ades an A/D converter 8 which contains a frame grabber circuit to capture the analog image being produced and convert it into a digital format for analysis. Of course, where imaging sensors producing a digital image output are used, there is no need for an additional A/D converter in the processor. The digitized image is stored temporarily in an image memory 10.
An array processor 12 utilizes a RISC co-processor for accelerated processing of algorithms on array type data files.
Custom software and firmware 14 contains the routines for applying algorithms to the raw digital image to extract specific data for analysis and comparison as will be described in greater detail below. rc~;gimg and control functions are also performed by the software.
An external storage and reader/wz=ter 16 is connected with _ _g_ the image processor 6 via an interface 18. The reader/writer is used to write and read digital infr~rmation from a card or disk medium containing spe~.:ific information such as a signature of the image being processed or compared.
A serial/parallel interface 20 i:~ used to provide an input and output between the positior. dete::tor, camera actuator, and other types of access control apparatus such as a communication link to a host database 22. The link is used to establish communicatior_ be~ween the image processor and a host computer for data base interrogation, statistic logging, and other administrative functions.
A keyboard 24 and keyboard interface 26 are used to input manual information to be stored and used with the image information. The keyboard is also used to communicate with the overall system for diagnostics.
The CPU 28 handles the input/cutput, housekeeping and communication functions. Most ef the algorithm routines are handled by the array processor. Finally, a power supply 30 is provided to power the image processor 6.
The enrollment process will be described with reference to Fig. 2. The individual to be enrolled removes his glasses, if any, since glass distorts the thermal image, and looks into the reflective surface of the. imaging sensor. To obtain a valid image, it is important that the subject be centered within the image frame.
The attendant then initiate' data entry 32 by activating the imaging sensor to acquir' an enalog thermal image 34 of the subject's face in accordance with the thermal energy patterns thereof. The image is tested for validity and clarity 36 and if insufficient, a substitute analog image is generated. An example of an analog thermal image of an individual is shown in Fig. 4.
The thermal image is converted into a digital format, a sample of which is shown in Fig. 5. The digital representation is essentially a t::atrix or array of pixels.
The intensity of each pixel currespc.~nds with the level of thermal energy of a corresponding p_acion of the analog .", ,: . ~, L:.. ,.'- _ CA~02083380 1999-04-28 thermal image. The digital raster data is stored 38 for subsequent normalization and feature extraction.
From the digital representation of the image, the edge surrounding the thermal image of the face is detected 40 using conventional edge detection techniques, Next, a vertical axis Y is created by locating a linE ~-hro~agh the centers of symmetry of four horizontal scan lines located in the center thirty percent of the imaaP. A secondary horizontal axis X1 is created by constructing a ~ine through the approximate center of the two hottest areas (i.e. those a=eas of greatest intensity), one on either side of thA nose bridge near the sinuses. This line is adjusted to be perpendicular to the vertical axis Y. Next, a primary horizontal axis X2 is generated by creating a horizontal line parallel to the secondary horizontal axis and offsetting the line down by a predetermined percent of the pixel c-~,znt of the horizontal axis.
A portion of the image -- referred to as a faceball -- is constructed 42 by forming the largest circle totally within the facial area. In general, there is only one center position for the largest circle. If more than one center position is located, then the most central position will be used as determined by comparing symmetrical features of the thermal signatures of the sinuses and cheeks of the individual. Once the faceball roundary is defined, all data related to thermal image portions ly-=~g outside the boundary is discarded to produce a non-ncrm~:liz~d faceball.
The faceball is normalized for size 4i via measuring and scaling its dimensions using bilinear interpolation in order to create a standard size (Fig. 6). The image is next scanned for minimum and maximum levels ever a predetermined rectangle encompassing features of the eye region, nose region and cheek region containing the elemental shapes. These measurements are used to generate a specific equalization correction transformation (feature enhancement) to achieve a standard histogram 46. This transformation i~ then applied to all the data in the faceball to create a new Equalized faceball as ,~:.
shown in Fig 7. The normalized facehall may be stored in a database and/or on an idelitizication card.
Analysis of the norm~_'_i~«ed fdceball is performed by feature extraction 48 wr~rei~: ele~ie:~cal shapes of the hottest and coldest thermal conto~~rs c:orresr,~onding to the sinuses, cheeks, and nose are ider_tifie~ Elw:uental shapes A, B, C, and D are shown in the image of Fig. 7.
An elemental shape is a contour in n-space,~the shape of which is related to the extent, intensity, and duration of output from a biosensor as rep_esented by an n-dimensional image. An elemental shape is the smallest replica of nested concentric contours centered at a position of interest. The location and contour of the elemental shape is determined by the physiological features detected by the biosensor. The size of the elemental shape represented in the image grows or shrinks as a function of parameters associated with the biosensor, procedures for its use, and the imaging system.
These parameters included:
a. sensitivity of the biosensor;
b. threshold settings for imac,e intensity levels;
c. distance between sensor and subject;
d. resolution or spot size of the biosensor;
e. blurring and damping effects of overlaying tissue;
f. repositioning movements during imaging;
g. intervening transducers and gels;
h. artifacts introduced by the imaging and display systems;
i. dynamic changes in the physiological system during imaging; and j. changes introduced by testing and treatment regimens.
The locations of interest depend upon the objective and the biosensor. In the case of thermal imaging for identification, the primary elemental shapes are located in the areas of the sinuses, the cheeks, and the nose. The elemental shape may be used as a fra~tal to perform compression of the data from the biosensor. By selecting ..
elemental shapes located at areas of greatest concern, image compression may be perform:W with negligible risk that the most essential information ,gill be lost through compression and decompression. Furtherr,~~~re, the elemental shape may be used to search a database of ,compressed images in order to locate a match between known and unknown subjects; to identify changes from time to time in a known subject; or to search for a known condition oceurr-r.g among subjects.
Through feature extraction of a normalized faceball, a signature of the individual may be defined and stored. Three different techniques may be used for feature extraction.
The preferred technique of feature extraction is to replicate bit maps of geometric areas known to contain the elemental shapes. Since the five areas containing features have been identified, trey can s~mr~y be cut out based upon a standard set of templates. These mapped areas are extracted along with information of where they were taken from on the original faceball and subsequently stored for later comparison.
A second technique is g~omeczic analysis wherein elemental shapes are identified and mapped as a shape vector table and stored with the location of it~~ centroid in the original faceball.
Finally, mathematical analysis of fractals may be used.
In this technique, a specific e'Unental shape is identified and stored as a fractal represFmt~.ti-.~n along with information relating to its location in the or.igir_al faceball.
Regardless of the feature extraction technique, the extracted data file is autocorrelated 50 and tested for validity 52 for accurate signature generation 54 of the individual. The data is then stored in a storage device or written on a facecard 56 to provide a permanent signature record.
The present invention is particularly suited for use in connection with security area access or entry protection through a recognition of aathorzzed -_=dividuals. Use of the invention for authorized acces:~ tv a sec~.zre area will be .~ ... r described with reference to Fig. 3.
An individual seeking access through a controlled checkpoint is detected at the entry door 58. The individual positions himself relative to an image sensor camera 60 and the positioning is checked 62 so that a full view of the face of the individual is obtained. The subject inserts a facecard bearing the subject signature into a data reader 64. The signature in the data acts as a reference for comparison with a new signature which is generated through the same steps 34-48 as described above in connection with Fig. 2.
Following feature extraction at step 48, correlation analysis 66 is performed to compare the reference data image from the card with the data derived from the contemporaneously generated image.
More particularly, for statistical analysis a cross correlation is performed in which an elemental shape of the reference faceball is correlated across an area of the .currently presented faceball. This correlation is represented in Fig. 8 with respect to the elemental shape B. The area is set to enclose all possible transpositions of that thermal feature which could be due to twisting, tilting or tipping of either faceball relative to the other or to variations in the imaging sensor.
A detailed representation of elemental shape B is shown in Fig. 9. The feature area is bounded by minimum and maximum values of coordinates a and b. The correlation matrix C (x, y) ~s derived from amax borax R(a, b~ - U(a-x, b-y>
- amen b - bm~n where R (a, b) is the reference faceball U (a, b) is the unknown current faceball -1 ~-If the resulting cross correlation matrix contains a global minimum, that minimum is compared to a predetermined threshold value representing a required quality of fit for recognition 68.
If the global minimum is below the threshold, the current faceball is recognized as being the same as the reference 70 and admission to the secure area is afforded 72. An entry report is made 74 and the system ret~irns to its idle state 76.
A good match is shown in Fig. 10.
If the global minimum is above the threshold, then other local minimums are evaluated. If the cross correlation matrix contains more than one local minimum, then each minimum is examined in order starting at the center of the matrix and proceeding outwardly. A local minimum is considered of valid interest if it is the center of a number of concentric contours of increasing valae. The tightness and number of the concentric contours is computed, i.e, the number of concentric rings occurring about the local minimum within the matrix. In cases where the local minimum occurs near the edge of the cross correlation matrix, incompleted rings are counted as if they were completed. A local minimum which does not have at least the minimum number of r-~~gs (set to two or more) is discarded, and the next minimum is considered. If a local minimum has at least the minimum number of rings, then it is compared to the threshold. If is is less, then the current faceball is recognized as being the reference or known person.
If too many comparisons are performed in which the global minimum is above the threshold 78, the system determines that the individual does not correspond with the reference from the identification card. A bad match is shown in Fig. 11. A
purge sequence is initiated 80, an entry denial report is issued 82, and the system returns to the idle state without granting access to the individual.
Rather than compare a current thermal image with one entered via an identification card, comparison may be made with a number of known faceballs stored in the system memory and corresponding with thermal images of authorized ... ,.~,",~
individuals.
If no global minimum i~ 1~:~at.ed which meets the threshold test of a first stored faceball, the unknown or currently generated faceball is matched against the next known faceball in the memory.
If additional security is required, several elemental shapes may be compared between :mown and unknown faces, before recognition is claimed. In the case comparing a presented face to the face represented on an identification card, one such pass through the comparisc~_~ algorithm yields the required match/no match decision. In the case where a presented face is to be matched against ~_ d=.~.ta base of known facts, it will speed the search if the data base is arranged according to similarly shaped elementary contours.
The elemental shapes may be considered in using a fractal representation of the known or unknown faces. If the database includes fractal-compressed face signatures, then the search may compare first the elemental shapes, and then the fractal-encoded patterns, ratter rn:~ requiring each candidate known face to be first decompressed pr'_.~r to correlation.
The threshold is see, for a particular hardware embodiment and application, in any adaptive cnanner such that it recognizes known faces and w.~en they have a reasonable amount of inconsistency in position and orientation, and in face and ambient temperature ranges ar_d :;~t does not mismatch unknown faces to similar known faces.
For geometric analysis, a faceball signature is read from a card that the individual inserts into a card reader. The data from the card is used as a reference for comparison to the newly acquired data. Al~ernatively, the reference signature may be accessed from a centralized data base by use of keyboard command, voice commaad, or other types of identifiers. The retrieved file is then appended to the faceball data file. The faceball area of search is defined by the location of the centroiC of the extracted original faceball shape ~ a predetermined percentage of the largest dimension both horizontally and vertically. The results of .,: i_ the comparison are then mapped to a separate results file. A
shape vector v~.!ap and pixel ~oc~tm~n ~f the centroid are saved in the file.
The resultant vector mad: and centroid are compared to the original signature file and r~?.e results are recorded. These results are then compared to a ~~redet:ermined set of threshold data used to determine the prob~biLity of match. Each feature is analyzed individually based on a rankir:g order and a decision to compare additional features is made based on the previous result. A single pass fail result is then issued.
Finally, for fractal analysis, the original signature file is either read from a card that the individual is carrying or the file may be accessed from a central data base by use of keyboard command, a voice command or other types of identifier. The retrieved fi'e is then appended to the faceball data file. For fracta' analysis, the faceball area of search is defined by the location of the extracted fractal in the original faceball to a predetermined percent of the expanded fractal shape in both vertical and horizontal dimensions. The results of the search and comparison are then mapped to a separate results file.
The resultant data ar_d location cf closest match are compared to the original signature file and the results are recorded. These results :_~re then compared to a predetermined set of threshold data used to d~=termine the probability of match. Each feature is ar..alyzPd individually based on a ranking order and a decision to compare additional features is made based on the previous result. A single pass fail result is then issued.
There has beer described and i7_lustrated herein a fully secure apparatus and method for positively and uniquely identifying an individual, an<' a security system utilizing the identificatior_. apparatus any? .-~~~thcd. While particular embodiments of the invention have been described, it is not intended that the inventi~~r. be limited thereby, as it is intended that the invention re oroad in scope and the specifications be read likewise. Thug, while a preferred ,..-.-.r.,..,....y..,.,, ,,...~~.~.,.~..~.., rv -1 ~-algorithm for identification has been described, the algorithm may be changed and refined in ma~v~ ways. Moreover, the criteria for determining that a ''match" has been found may be set according to various requirements. Thus, depending on the level of security desired, the storage capacity of the card, and the accuracy of the system provic.~ed, a match may be determined to have been found -f one or more elemental shape is similar.
While particular elemental shapes and quality of fit criteria were described as being preferred for determining an individual's identity, those skilled in the art will recognize that depending upon the desired accuracy of the system, fewer or more parameters, as well as different parameters could be utilized. Likewise, while discussion was had with regard to obtaining facial image informa~ion from a frontal view with a scanning thermal imager sensitive in the 3 to 6 micron range, it will be appreciated that information could also be gleaned from other types of thermal :ma~~ers, at other wavelengths, as well as from other types of ~iosensors whose output can be represented as n-dimensional images. Indeed the techniques of the invention can be used to dentify animals or abnormalities in the body such as migraine readaches, inflammation, arthritis, arterial blockages, or the presence of false facial hair.
While particular biosensor data has been described, those skilled in the art will recognize that depending upon the desired accuracy of the system, fewer or more biosensors, as well as different biosenscr data ~.~resenta.~ions could be utilized. Also, whip the storing of reference sensor data is preferably in the form of grey scale recordings of 8 bit data from an array of at least. 256 x 240 pixels, indications of values may be obtained ar_d stored in many ways. For example, the information storage means or_ the identification card could range, e.g., from a lamir_ated magnetic strip, to an embedded integrated circuit, tc; encoding by laser deformation on a metallic card.
Further yet, it will be 2r~rP'ia~::ed that the computing _1')_ equipment could vary greatly, as different types of computers and data entry devices could k~A used. Various techniques could also be employed for arranging that the face to be identified is aligned witr~ ~:e imaging biosensor, and various correlation techniques gar. be employed for searching a data file for the best match or comparing images resulting from a presented card and from its holder. Therefore, it will be apparent to those skilled in the art that many other changes and modifications may be made to the invention as described in the specification without departi_~g from the spirit and scope of the invention as so claimed.
OF ELEMENTAL SHAPES DERIVED FROM BIOSENSOR DATA
BACKGROUND OF THE INVENTION
The problem of ascertaining the identity of an individual or verifying whether an individual is the person he claims to be is a common one continually fa.ed by individuals, businesses, and governments Me'; od for positive identification of an individual usually include reliance upon knowledge of restricted information (such as a password), possession of a restricted article (such as a passkey), or physical appearance (such as matching a reference photo).
Security based upon knowledge or possession may be compromised without discovery since r_he information or article may be extorted from its rightful owner. The third methodology, commonly referred to as biometric techniques, is considered less vulnerable to m~:~take.n identity. The best known biometric technique .~md tr~F:: one considered most secure is fingerprint analysis. Hou~Pver, other biometric techniques exist. The present inventyor~ relates to an identification technique utilizing thermal imaging.
Skin temperature is a result of the balance between the heat flow from tissue to the skin surface and the heat flow from the surface to the surroundings. The basic mechanisms for transport of heat from the tissues to the skin surface include conduction throughout the tissues and convection by the flow of blood from the warm art:~rial supply through the capillary bed into the venous plex~~s. Metabolic heat generation is a second order influence compared to the convection transport mechanism of the blood perfu.,ion. Thus in the absence of active superficial veins, the diffuse avascular skin temperature is directly related to the blood perfusion per unit volume of tissue, including the skin itself as well as tissues approximately 2 to 3 em deep to the skin. Active superficial veins are also a possible indicator of regional blood flow. These azn dr..~~mstrat~d as well-defined patterns of increased skin temperature. Skin temperature immediately overlying the vein is only a few tenths of a degree Celsius lower than the temperature of the blood flowing in the vein itself .
A variation of 3.5 degrees Fahrenheit is usual across a human facial surface. In the ::ase of severe medical abnormalities, or when the ambient temperature is severe, the range may be twice as large, and its average value may vary.
The heat energy given off by the face is changed to electrical energy and greatly amplified by the electronic circuitry of a thermal imaging system. Following amplification, the electrical energy is digitized. A facial "thermogram" is the resultant two-dimensional image obtained by scanning each section of the face onto one or more detectors which are sensitive in the infrared range. The most commonly used medical thermal imagers produce an image which comprises approximately 60,000 individual heat recordings. Most commonly used scanning infrared cameras do not accurately measure vein temperature because the modulation transfer function of the optics of these cameras produces a measurement that is related to the average temperature over almost 1 cm2 of scanning surface. Thus the size of the vein will influence the indicated temperature.
Because a thermograph operates at a distance from the subject and detects and records only radiant heat spontaneously emitted from the body surface, it constitutes a painless, non-invasive, passive method of recording body surface temperatures. In general, the nose, cheeks, ears, and the avascular corneas of the ey:s are cool, as are the hair-insulated eyebrow, eyelash, and scalp regions. The anterior portion of the forehead is warm and has a narrow range of temperature variation over its surface. Heat is normally trapped by skin folds, at the canthi and along the free margins of the eyelids. EssentiGlly symmetrical facial thermograms are found almost without exception in healthy subjects.
In accordance with the ~.~resent invention, facial thermograms are processed to extract elemental shapes therefrom. the elemental shapes -- like a fingerprint --are unique to an individual. Thus, the identity of an individual may be asce---tamed by comparing current elemental shapes with those from a z~~erence.
BRIEF DESCRIPTION OF THE PRIOR ART
Individuals identify one ar_other primarily by recognizing their faces. It is therefore understandable that systems would be developed which would attempt to replicate how persons recognize one another. Various patents have been issued which utilize known photographic or other visual images for comparison with unknown imagea in order to automate the recognition task. In the facial curve method of the Rothfjell U.S. Patent No. 3,805,238, facial curves (two or three dimensional) of an individual t~~ken from fullface, forty-five degree angle, and profile (ninety degree angle) are stored on an identity card or in a data processing machine.
Identification of the individual is performed by first visually comparing the photographs on the identity card to the individual presenting the card and then taking an image of the individual and automatically comparing by computer the curves stored on the identify card with curves derived from the image. In order to properly compare such an image with the computer data, however, the individual must be positioned and aligned in front of a glass screen.
It is also known in the art to measure facial features utilizing ratios obtained from a set of measurements taken from a facial image as disclosed in the Tal U.S. patent No.
4,975,969. The use of binary images obtained by thresholding facial images collected under relatively uncontrolled conditions and finding the best :r,atch among small groups of known indi~=iduals is disclosed i_~ the U.S. patent to Lu No.
4,858,000.
The time ~equ~ -~ed fear enrol-i mer_'~ and verification, the difficulty in automat-i.r:g the tec}rniques, and variabilities resulting from lighting conditions and shadows are major Y:~;? ~.
_ ~Y
limitations to the ro~.aine, unattended use of these and other facial recognition systems whicY_~~tilize photographs or video images. In addition, sucr~ systems are vulnerable to forgery through disguise and use ~f photographs or mannequins.
Identification tecr~niques utilizing infrared thermal imaging are known in the art as evidenced by the U.S. patent to Rice No. 4,699,149. This pate~~c is predicated upon the discovery that the arrangement _- veins in a local portion of an individual's body is a uniiue characteristic of that individual. IdentificatioYi is pe-formed by linearly scanning a region of skin at the wris, and detecting the position of subcutaneous blood vessels by measuring a parameter such as the reflection of incident radiation. A transducer is constrained to follow a particular path, and the resulting linear scan measurements arE thresholded to produce a binary image which constitutes the identifs:ation signature. The measured signature and predetexo,inec~ signature at the same location are compared.
A number of drawbacks are inherent in the Rice system.
for example, active illumination cY heating of the area being scanned is required, as is identical replication of the small skin area selected for scanning. These requirements may cause discomfort to the individual. Moreover, identification can not be verified from a distance as is necessary for security access systems.
The present invention was de~reloped in order to overcome these and other drawbacks of the prior identification techniques by providing a passi~-~ identification method wherein unique elemental shapes are extracted from a thermal image of the individual.
SUMMARY OF THE INVENTION
Accordingly, it is a prirrrary object of the present invention to provide a method for identifying individuals from biosensor data. A thermal image of a portion of the individual's body is generateu in accordance with the thermal energy patterns thereof. T:~e therm~:_ image is processed to ;: ~ ~'~ ° .., produce a digital representation thereof. The digital representation includes a ma rix of pixels and the intensity of each pixel correspor_~~s w~' ~ ~ tr-:.e lewal of thermal energy of a corresponding portior:.~f ne image. The digital image is normalized in order to elim_~.~_ate variables and produce a normalized image containing elemental shapes comprising contours corresponding with unique structural features of the individual. Normalization is accomplished by cropping the thermal image to a cropped area of maximum uniqueness of an individual relative to others, sucr~ as the eye and nose area of the individual°s face, stand~.rd~zing the image for size, and equalizing the image to a standard histogram. The processed image is stored for subsequent identification of the individual.
According to another object or the invention, a region of the image containing specific elemental shapes of interest is defined by geometry, mathematics, or as a bit-map replica of the region in order to provide ~_ signature of the individual.
Correlation techniques are used to compare a previously defined reference signature vi an individual with a currently defined signature to cor:fir.; the individual's identify.
BRIEF DESCRIPTION OF THE FIGJRES
Other objects and advantages of the invention will become apparent from a st~~dy of the fol~_r~wing specification when viewed in the light of the ac~~ompar_ying drawing, in which:
Fig. 1 is a block diagram of the apparatus used to create an identification signature of ar! individual;
Fig. 2 is a flow chart illustrating the steps for creating the identification signature of an individual from biosensor data in accordance with the invention;
Fig. 3 is a flow chart illustrating the steps for identifying an individual to_ascertain whether entry of the individual to a secure area is to be provided;
Fig. 4 is a digitized thermal image of an individual's face;
Fig. 5 is the image of Fig. 4 including axes used for ' r . :v,~. z s ..
-G-cropping;
Fig. 6 is a digitized th:~rmal contcur image of the image of Fig. 4 after normalizatic~z for size;
Fig. 7 is the image of F~e~. 5 following histogram equalization;
Fig. 8 illustrates correlat~om of two images to identify a match of certain features thereof;
Fig. 9 is a detailed illast._atir~n of do elemental shape;
and Figs. 10 and 11 ill~_zst~ate the correlation results of good and bad match areas, respectively.
DETAILED DESCRIPTION
The invention relates to ~. method and apparatus for uniquely identifying individu_.~s through analysis of elemental shapes derived from biosensor data p-cesented in n dimensions.
In order to conform the identity of an individual, a reference identification for that individual must be obtained.
This is performed by enrol~.ing or storing a thermogram image of the individual in a storage ~?ev~ce such as a storage card or a memory for subsequent comparison. In this manner, when the individual presents himself for identification, such as when he desires to gain entry to a secure area, the current thermogram image of the in,~ivi~.ual can be compared with the stored reference to confirm the individual's identity. the apparatus for generatir_g a thermogram image of the individual will be described with reference to Fig. 1.
A position detector 2 is ~~se.to determine whether the individual is properly framed ir~ front of an image sensor 4.
In the preferred embodiment, the face of the individual is that portion positioned in front of the sensor from which the thermogram is generated. As used herein the term face includes all portions c~f t.ne head imaged by the sensor 4. The degrees of tilt, twist, and tip can be made with reference to polar coordinates establishing ~'ie ''faceball" as will be described below. The preferred emboc~imemt results in a standard 30° upwards angle t~ th= fa~:e during imaging, which reduces the possibility of impinging hair or clothing on the minimum portion of the f~:ce r~eauired for most accurate identification, that p~~ri io:_ ~..eing the swath occurring between the approximately noriz~mcal 1-nes established by the uppermost points of the eyebrows and by the nostrils or lowest portion of the nose.
The image sensor 4 preferably comprises a platinum silicide staring array camera although a ~-ariety of other infrared imagers of e-they the array or scanning type may be used. Furthermore, other t:.~pes ~n n-dimensional biometric sensors may be used including for example magnetic resonance imagers, digital angiographs, computed tomographs, sonographs, radioisotope imagers, and X-_~ay devices. Infrared imagers producing emissions in the 3-6 or 8-14 micron ranges are preferred since thermal data from the individual may be collected passively and instantaneously with no risk to the individual.
The image sensor 4 produces an analog image of the individual's face wherein the intensity level is direction proportional to the absolute temperature at points across the face, where the size of the points depends on the imager used.
The analog image is sent to ar_ image processor 6 for analysis. The processor inc~ades an A/D converter 8 which contains a frame grabber circuit to capture the analog image being produced and convert it into a digital format for analysis. Of course, where imaging sensors producing a digital image output are used, there is no need for an additional A/D converter in the processor. The digitized image is stored temporarily in an image memory 10.
An array processor 12 utilizes a RISC co-processor for accelerated processing of algorithms on array type data files.
Custom software and firmware 14 contains the routines for applying algorithms to the raw digital image to extract specific data for analysis and comparison as will be described in greater detail below. rc~;gimg and control functions are also performed by the software.
An external storage and reader/wz=ter 16 is connected with _ _g_ the image processor 6 via an interface 18. The reader/writer is used to write and read digital infr~rmation from a card or disk medium containing spe~.:ific information such as a signature of the image being processed or compared.
A serial/parallel interface 20 i:~ used to provide an input and output between the positior. dete::tor, camera actuator, and other types of access control apparatus such as a communication link to a host database 22. The link is used to establish communicatior_ be~ween the image processor and a host computer for data base interrogation, statistic logging, and other administrative functions.
A keyboard 24 and keyboard interface 26 are used to input manual information to be stored and used with the image information. The keyboard is also used to communicate with the overall system for diagnostics.
The CPU 28 handles the input/cutput, housekeeping and communication functions. Most ef the algorithm routines are handled by the array processor. Finally, a power supply 30 is provided to power the image processor 6.
The enrollment process will be described with reference to Fig. 2. The individual to be enrolled removes his glasses, if any, since glass distorts the thermal image, and looks into the reflective surface of the. imaging sensor. To obtain a valid image, it is important that the subject be centered within the image frame.
The attendant then initiate' data entry 32 by activating the imaging sensor to acquir' an enalog thermal image 34 of the subject's face in accordance with the thermal energy patterns thereof. The image is tested for validity and clarity 36 and if insufficient, a substitute analog image is generated. An example of an analog thermal image of an individual is shown in Fig. 4.
The thermal image is converted into a digital format, a sample of which is shown in Fig. 5. The digital representation is essentially a t::atrix or array of pixels.
The intensity of each pixel currespc.~nds with the level of thermal energy of a corresponding p_acion of the analog .", ,: . ~, L:.. ,.'- _ CA~02083380 1999-04-28 thermal image. The digital raster data is stored 38 for subsequent normalization and feature extraction.
From the digital representation of the image, the edge surrounding the thermal image of the face is detected 40 using conventional edge detection techniques, Next, a vertical axis Y is created by locating a linE ~-hro~agh the centers of symmetry of four horizontal scan lines located in the center thirty percent of the imaaP. A secondary horizontal axis X1 is created by constructing a ~ine through the approximate center of the two hottest areas (i.e. those a=eas of greatest intensity), one on either side of thA nose bridge near the sinuses. This line is adjusted to be perpendicular to the vertical axis Y. Next, a primary horizontal axis X2 is generated by creating a horizontal line parallel to the secondary horizontal axis and offsetting the line down by a predetermined percent of the pixel c-~,znt of the horizontal axis.
A portion of the image -- referred to as a faceball -- is constructed 42 by forming the largest circle totally within the facial area. In general, there is only one center position for the largest circle. If more than one center position is located, then the most central position will be used as determined by comparing symmetrical features of the thermal signatures of the sinuses and cheeks of the individual. Once the faceball roundary is defined, all data related to thermal image portions ly-=~g outside the boundary is discarded to produce a non-ncrm~:liz~d faceball.
The faceball is normalized for size 4i via measuring and scaling its dimensions using bilinear interpolation in order to create a standard size (Fig. 6). The image is next scanned for minimum and maximum levels ever a predetermined rectangle encompassing features of the eye region, nose region and cheek region containing the elemental shapes. These measurements are used to generate a specific equalization correction transformation (feature enhancement) to achieve a standard histogram 46. This transformation i~ then applied to all the data in the faceball to create a new Equalized faceball as ,~:.
shown in Fig 7. The normalized facehall may be stored in a database and/or on an idelitizication card.
Analysis of the norm~_'_i~«ed fdceball is performed by feature extraction 48 wr~rei~: ele~ie:~cal shapes of the hottest and coldest thermal conto~~rs c:orresr,~onding to the sinuses, cheeks, and nose are ider_tifie~ Elw:uental shapes A, B, C, and D are shown in the image of Fig. 7.
An elemental shape is a contour in n-space,~the shape of which is related to the extent, intensity, and duration of output from a biosensor as rep_esented by an n-dimensional image. An elemental shape is the smallest replica of nested concentric contours centered at a position of interest. The location and contour of the elemental shape is determined by the physiological features detected by the biosensor. The size of the elemental shape represented in the image grows or shrinks as a function of parameters associated with the biosensor, procedures for its use, and the imaging system.
These parameters included:
a. sensitivity of the biosensor;
b. threshold settings for imac,e intensity levels;
c. distance between sensor and subject;
d. resolution or spot size of the biosensor;
e. blurring and damping effects of overlaying tissue;
f. repositioning movements during imaging;
g. intervening transducers and gels;
h. artifacts introduced by the imaging and display systems;
i. dynamic changes in the physiological system during imaging; and j. changes introduced by testing and treatment regimens.
The locations of interest depend upon the objective and the biosensor. In the case of thermal imaging for identification, the primary elemental shapes are located in the areas of the sinuses, the cheeks, and the nose. The elemental shape may be used as a fra~tal to perform compression of the data from the biosensor. By selecting ..
elemental shapes located at areas of greatest concern, image compression may be perform:W with negligible risk that the most essential information ,gill be lost through compression and decompression. Furtherr,~~~re, the elemental shape may be used to search a database of ,compressed images in order to locate a match between known and unknown subjects; to identify changes from time to time in a known subject; or to search for a known condition oceurr-r.g among subjects.
Through feature extraction of a normalized faceball, a signature of the individual may be defined and stored. Three different techniques may be used for feature extraction.
The preferred technique of feature extraction is to replicate bit maps of geometric areas known to contain the elemental shapes. Since the five areas containing features have been identified, trey can s~mr~y be cut out based upon a standard set of templates. These mapped areas are extracted along with information of where they were taken from on the original faceball and subsequently stored for later comparison.
A second technique is g~omeczic analysis wherein elemental shapes are identified and mapped as a shape vector table and stored with the location of it~~ centroid in the original faceball.
Finally, mathematical analysis of fractals may be used.
In this technique, a specific e'Unental shape is identified and stored as a fractal represFmt~.ti-.~n along with information relating to its location in the or.igir_al faceball.
Regardless of the feature extraction technique, the extracted data file is autocorrelated 50 and tested for validity 52 for accurate signature generation 54 of the individual. The data is then stored in a storage device or written on a facecard 56 to provide a permanent signature record.
The present invention is particularly suited for use in connection with security area access or entry protection through a recognition of aathorzzed -_=dividuals. Use of the invention for authorized acces:~ tv a sec~.zre area will be .~ ... r described with reference to Fig. 3.
An individual seeking access through a controlled checkpoint is detected at the entry door 58. The individual positions himself relative to an image sensor camera 60 and the positioning is checked 62 so that a full view of the face of the individual is obtained. The subject inserts a facecard bearing the subject signature into a data reader 64. The signature in the data acts as a reference for comparison with a new signature which is generated through the same steps 34-48 as described above in connection with Fig. 2.
Following feature extraction at step 48, correlation analysis 66 is performed to compare the reference data image from the card with the data derived from the contemporaneously generated image.
More particularly, for statistical analysis a cross correlation is performed in which an elemental shape of the reference faceball is correlated across an area of the .currently presented faceball. This correlation is represented in Fig. 8 with respect to the elemental shape B. The area is set to enclose all possible transpositions of that thermal feature which could be due to twisting, tilting or tipping of either faceball relative to the other or to variations in the imaging sensor.
A detailed representation of elemental shape B is shown in Fig. 9. The feature area is bounded by minimum and maximum values of coordinates a and b. The correlation matrix C (x, y) ~s derived from amax borax R(a, b~ - U(a-x, b-y>
- amen b - bm~n where R (a, b) is the reference faceball U (a, b) is the unknown current faceball -1 ~-If the resulting cross correlation matrix contains a global minimum, that minimum is compared to a predetermined threshold value representing a required quality of fit for recognition 68.
If the global minimum is below the threshold, the current faceball is recognized as being the same as the reference 70 and admission to the secure area is afforded 72. An entry report is made 74 and the system ret~irns to its idle state 76.
A good match is shown in Fig. 10.
If the global minimum is above the threshold, then other local minimums are evaluated. If the cross correlation matrix contains more than one local minimum, then each minimum is examined in order starting at the center of the matrix and proceeding outwardly. A local minimum is considered of valid interest if it is the center of a number of concentric contours of increasing valae. The tightness and number of the concentric contours is computed, i.e, the number of concentric rings occurring about the local minimum within the matrix. In cases where the local minimum occurs near the edge of the cross correlation matrix, incompleted rings are counted as if they were completed. A local minimum which does not have at least the minimum number of r-~~gs (set to two or more) is discarded, and the next minimum is considered. If a local minimum has at least the minimum number of rings, then it is compared to the threshold. If is is less, then the current faceball is recognized as being the reference or known person.
If too many comparisons are performed in which the global minimum is above the threshold 78, the system determines that the individual does not correspond with the reference from the identification card. A bad match is shown in Fig. 11. A
purge sequence is initiated 80, an entry denial report is issued 82, and the system returns to the idle state without granting access to the individual.
Rather than compare a current thermal image with one entered via an identification card, comparison may be made with a number of known faceballs stored in the system memory and corresponding with thermal images of authorized ... ,.~,",~
individuals.
If no global minimum i~ 1~:~at.ed which meets the threshold test of a first stored faceball, the unknown or currently generated faceball is matched against the next known faceball in the memory.
If additional security is required, several elemental shapes may be compared between :mown and unknown faces, before recognition is claimed. In the case comparing a presented face to the face represented on an identification card, one such pass through the comparisc~_~ algorithm yields the required match/no match decision. In the case where a presented face is to be matched against ~_ d=.~.ta base of known facts, it will speed the search if the data base is arranged according to similarly shaped elementary contours.
The elemental shapes may be considered in using a fractal representation of the known or unknown faces. If the database includes fractal-compressed face signatures, then the search may compare first the elemental shapes, and then the fractal-encoded patterns, ratter rn:~ requiring each candidate known face to be first decompressed pr'_.~r to correlation.
The threshold is see, for a particular hardware embodiment and application, in any adaptive cnanner such that it recognizes known faces and w.~en they have a reasonable amount of inconsistency in position and orientation, and in face and ambient temperature ranges ar_d :;~t does not mismatch unknown faces to similar known faces.
For geometric analysis, a faceball signature is read from a card that the individual inserts into a card reader. The data from the card is used as a reference for comparison to the newly acquired data. Al~ernatively, the reference signature may be accessed from a centralized data base by use of keyboard command, voice commaad, or other types of identifiers. The retrieved file is then appended to the faceball data file. The faceball area of search is defined by the location of the centroiC of the extracted original faceball shape ~ a predetermined percentage of the largest dimension both horizontally and vertically. The results of .,: i_ the comparison are then mapped to a separate results file. A
shape vector v~.!ap and pixel ~oc~tm~n ~f the centroid are saved in the file.
The resultant vector mad: and centroid are compared to the original signature file and r~?.e results are recorded. These results are then compared to a ~~redet:ermined set of threshold data used to determine the prob~biLity of match. Each feature is analyzed individually based on a rankir:g order and a decision to compare additional features is made based on the previous result. A single pass fail result is then issued.
Finally, for fractal analysis, the original signature file is either read from a card that the individual is carrying or the file may be accessed from a central data base by use of keyboard command, a voice command or other types of identifier. The retrieved fi'e is then appended to the faceball data file. For fracta' analysis, the faceball area of search is defined by the location of the extracted fractal in the original faceball to a predetermined percent of the expanded fractal shape in both vertical and horizontal dimensions. The results of the search and comparison are then mapped to a separate results file.
The resultant data ar_d location cf closest match are compared to the original signature file and the results are recorded. These results :_~re then compared to a predetermined set of threshold data used to d~=termine the probability of match. Each feature is ar..alyzPd individually based on a ranking order and a decision to compare additional features is made based on the previous result. A single pass fail result is then issued.
There has beer described and i7_lustrated herein a fully secure apparatus and method for positively and uniquely identifying an individual, an<' a security system utilizing the identificatior_. apparatus any? .-~~~thcd. While particular embodiments of the invention have been described, it is not intended that the inventi~~r. be limited thereby, as it is intended that the invention re oroad in scope and the specifications be read likewise. Thug, while a preferred ,..-.-.r.,..,....y..,.,, ,,...~~.~.,.~..~.., rv -1 ~-algorithm for identification has been described, the algorithm may be changed and refined in ma~v~ ways. Moreover, the criteria for determining that a ''match" has been found may be set according to various requirements. Thus, depending on the level of security desired, the storage capacity of the card, and the accuracy of the system provic.~ed, a match may be determined to have been found -f one or more elemental shape is similar.
While particular elemental shapes and quality of fit criteria were described as being preferred for determining an individual's identity, those skilled in the art will recognize that depending upon the desired accuracy of the system, fewer or more parameters, as well as different parameters could be utilized. Likewise, while discussion was had with regard to obtaining facial image informa~ion from a frontal view with a scanning thermal imager sensitive in the 3 to 6 micron range, it will be appreciated that information could also be gleaned from other types of thermal :ma~~ers, at other wavelengths, as well as from other types of ~iosensors whose output can be represented as n-dimensional images. Indeed the techniques of the invention can be used to dentify animals or abnormalities in the body such as migraine readaches, inflammation, arthritis, arterial blockages, or the presence of false facial hair.
While particular biosensor data has been described, those skilled in the art will recognize that depending upon the desired accuracy of the system, fewer or more biosensors, as well as different biosenscr data ~.~resenta.~ions could be utilized. Also, whip the storing of reference sensor data is preferably in the form of grey scale recordings of 8 bit data from an array of at least. 256 x 240 pixels, indications of values may be obtained ar_d stored in many ways. For example, the information storage means or_ the identification card could range, e.g., from a lamir_ated magnetic strip, to an embedded integrated circuit, tc; encoding by laser deformation on a metallic card.
Further yet, it will be 2r~rP'ia~::ed that the computing _1')_ equipment could vary greatly, as different types of computers and data entry devices could k~A used. Various techniques could also be employed for arranging that the face to be identified is aligned witr~ ~:e imaging biosensor, and various correlation techniques gar. be employed for searching a data file for the best match or comparing images resulting from a presented card and from its holder. Therefore, it will be apparent to those skilled in the art that many other changes and modifications may be made to the invention as described in the specification without departi_~g from the spirit and scope of the invention as so claimed.
Claims (16)
- A method for identifying individuals from biosensor data, comprising the steps of (a) generating a thermal image of a portion of an individual's body in accordance with thermal energy patterns thereof;
(b) processing said thermal image to produce a digital representation thereof, said digital representation comprising a matrix of pixels, each pixel having an intensity corresponding with the level of thermal energy of a corresponding portion of said image;
(c) normalizing said thermal image in order to eliminate variables therefrom, thereby to produce a normalized image containing elemental shapes comprising contours corresponding with unique structural features of the individual; and (d) producing and storing a digital representation of said normalized image, whereby an initial unique digital representation is created and stored for the individual as a reference representation for subsequent identification, through a comparison with a current unique digital representation for said individual created thereby. - 2. A method as defined in claim 1, wherein said normalizing step includes cropping said thermal image to obtain a cropped area therefrom, having maximum uniqueness of said structural features of the individual.
- 3. A method as defined in claim 2, wherein said cropped image area comprises a portion of the individual's face containing eyes and nose.
- 4. A method as defined in claim 2, wherein said cropping step is performed by (a) detecting an edge of said thermal image;
(b) defining a vertical axis of said thermal image within said edge by locating centres of symmetry of intensity levels along a horizontal line of pixels;
(c) defining a horizontal axis of said thermal image within said edge by locating approximate centres of areas of greatest intensity on either side of the vertical axis;
(d) defining a largest boundary enclosing said thermal image; and (e) deleting portions of said thermal image, lying outside said boundary. - 5. A method as defined in claim 2, wherein said normalizing step further includes the step of standardizing said thermal image for size by measuring dimensions of said cropped area and scaling said dimensions to preserve and reconstruct said thermal image to a standard size.
- 6. A method as defined in claim 5, wherein said normalizing step further includes the step of equalizing said thermal image to a standard histogram using minimum and maximum intensity levels over specific feature areas of said thermal image to generate said standard histogram.
- 7. A method as defined in claim 1, further comprising the step of identifying regions of said thermal image containing elemental shapes of interest.
- 8. A method as defined in claim 7, further comprising the step of extracting a feature including at least a portion of each region containing an elemental shape together with information on a location thereof within the thermal image, said feature to define a signature of the individual.
- 9. A method as defined in claim 8, wherein each of said regions is defined by mapping the elemental shapes as a shape vector table.
- 10. A method as defined in claim 8, wherein each of said regions is defined by replicating bit maps of geometric areas containing the elemental shapes.
- 11. A method as defined in claim 8, wherein each of said regions is defined by providing a fractal representation of the elemental shapes.
- 12. A method as defined in claim 8, and further comprising the steps of storing an initially defined signature for the individual as a reference signature, and comparing said reference signature with a currently defined signature for the individual to confirm the individual's identity.
- 13. A method as defined in claim 12, wherein said comparison step is accomplished using correlation techniques.
- 14. A method as defined in claim 12, wherein said comparison step comprises matching a location and a representation of each of said elemental shapes.
- 15. A method as defined in claim 14, further comprising the step of adaptively establishing a threshold level of match to confirm the individual's identity, whereby the threshold level must be met in order to confirm the individual's identity.
- 16. A method as defined in claim 15, wherein said portion of the individual's body comprises the face, and further wherein said normalized image contains elemental shapes corresponding with at least the eyes and nose of the individual.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US672,320 | 1991-03-20 | ||
| US07/672,320 US5163094A (en) | 1991-03-20 | 1991-03-20 | Method for identifying individuals from analysis of elemental shapes derived from biosensor data |
| PCT/US1992/001805 WO1992016910A1 (en) | 1991-03-20 | 1992-03-18 | Method for identifying individuals from analysis of elemental shapes derived from biosensor data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CA2083380A1 CA2083380A1 (en) | 1992-09-21 |
| CA2083380C true CA2083380C (en) | 1999-08-10 |
Family
ID=24698061
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA002083380A Expired - Fee Related CA2083380C (en) | 1991-03-20 | 1992-03-18 | Method for identifying individuals from analysis of elemental shapes derived from biosensor data |
Country Status (6)
| Country | Link |
|---|---|
| US (1) | US5163094A (en) |
| EP (1) | EP0533891B1 (en) |
| JP (1) | JP2873740B2 (en) |
| CA (1) | CA2083380C (en) |
| DE (1) | DE69232024T2 (en) |
| WO (1) | WO1992016910A1 (en) |
Families Citing this family (109)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10361802B1 (en) | 1999-02-01 | 2019-07-23 | Blanding Hovenweep, Llc | Adaptive pattern recognition based control system and method |
| JP2973676B2 (en) * | 1992-01-23 | 1999-11-08 | 松下電器産業株式会社 | Face image feature point extraction device |
| GB2264779B (en) * | 1992-02-20 | 1996-05-01 | Thermoteknix Systems Ltd | Monitoring changes in image characteristics |
| JP3252381B2 (en) * | 1992-09-08 | 2002-02-04 | ソニー株式会社 | Pattern recognition device |
| US5351303A (en) * | 1993-02-25 | 1994-09-27 | Willmore Michael R | Infra-red imaging and pattern recognition system |
| CA2125337A1 (en) * | 1993-06-30 | 1994-12-31 | Marlin Jay Eller | Method and system for searching compressed data |
| US5574573A (en) * | 1993-10-29 | 1996-11-12 | Eastman Kodak Company | Compression method for a standardized image library |
| US5835616A (en) * | 1994-02-18 | 1998-11-10 | University Of Central Florida | Face detection using templates |
| US5781650A (en) * | 1994-02-18 | 1998-07-14 | University Of Central Florida | Automatic feature detection and age classification of human faces in digital images |
| US5474085A (en) * | 1994-02-24 | 1995-12-12 | University Of Prince Edward Island | Remote thermographic sensing of livestock |
| JPH09510636A (en) * | 1994-03-24 | 1997-10-28 | ミネソタ マイニング アンド マニュファクチャリング カンパニー | Biometric personal identification system |
| US5497430A (en) * | 1994-11-07 | 1996-03-05 | Physical Optics Corporation | Method and apparatus for image recognition using invariant feature signals |
| US7152045B2 (en) * | 1994-11-28 | 2006-12-19 | Indivos Corporation | Tokenless identification system for authorization of electronic transactions and electronic transmissions |
| FR2736454B1 (en) * | 1995-07-03 | 1997-08-08 | Commissariat Energie Atomique | METHOD FOR RECONSTRUCTING THREE-DIMENSIONAL IMAGES ON A MOBILE OR DEFORMABLE OBJECT |
| JPH0991434A (en) * | 1995-09-28 | 1997-04-04 | Hamamatsu Photonics Kk | Human body collation device |
| US5892838A (en) * | 1996-06-11 | 1999-04-06 | Minnesota Mining And Manufacturing Company | Biometric recognition using a classification neural network |
| US6173068B1 (en) | 1996-07-29 | 2001-01-09 | Mikos, Ltd. | Method and apparatus for recognizing and classifying individuals based on minutiae |
| US5991429A (en) * | 1996-12-06 | 1999-11-23 | Coffin; Jeffrey S. | Facial recognition system for security access and identification |
| US5768647A (en) * | 1996-12-06 | 1998-06-16 | Technology Recognition Systems, Inc. | Camera positioning assembly |
| DE19700811A1 (en) * | 1997-01-13 | 1998-07-16 | Heinrich Landert | Method and device for controlling door systems depending on the presence of people |
| US6145738A (en) | 1997-02-06 | 2000-11-14 | Mr. Payroll Corporation | Method and apparatus for automatic check cashing |
| US6149056A (en) | 1997-02-06 | 2000-11-21 | Mr. Payroll Corporation | Automatic check cashing using biometric identification verification |
| JPH10322686A (en) * | 1997-05-15 | 1998-12-04 | Niles Parts Co Ltd | Door camera unit provided with image memory |
| US6560352B2 (en) * | 1999-10-08 | 2003-05-06 | Lumidigm, Inc. | Apparatus and method of biometric identification or verification of individuals using optical spectroscopy |
| US7890158B2 (en) | 2001-06-05 | 2011-02-15 | Lumidigm, Inc. | Apparatus and method of biometric determination using specialized optical spectroscopy systems |
| US6628809B1 (en) | 1999-10-08 | 2003-09-30 | Lumidigm, Inc. | Apparatus and method for identification of individuals by near-infrared spectrum |
| JP3436473B2 (en) * | 1997-06-20 | 2003-08-11 | シャープ株式会社 | Image processing device |
| US6026188A (en) * | 1997-10-10 | 2000-02-15 | Unisys Corporation | System and method for recognizing a 3-D object by generating a rotated 2-D image of the object from a set of 2-D enrollment images |
| US6185316B1 (en) | 1997-11-12 | 2001-02-06 | Unisys Corporation | Self-authentication apparatus and method |
| US6178259B1 (en) * | 1998-04-17 | 2001-01-23 | Unisys Corporation | Method of compensating for pixel histogram distortion in an object recognition system |
| WO1999057683A1 (en) | 1998-05-04 | 1999-11-11 | The Johns Hopkins University | Method and apparatus for segmenting small structures in images |
| US6343140B1 (en) | 1998-09-11 | 2002-01-29 | Quid Technologies Llc | Method and apparatus for shooting using biometric recognition |
| US6496594B1 (en) | 1998-10-22 | 2002-12-17 | Francine J. Prokoski | Method and apparatus for aligning and comparing images of the face and body from different imagers |
| US7305562B1 (en) | 1999-03-09 | 2007-12-04 | Citibank, N.A. | System, method and computer program product for an authentication management infrastructure |
| US6256737B1 (en) | 1999-03-09 | 2001-07-03 | Bionetrix Systems Corporation | System, method and computer program product for allowing access to enterprise resources using biometric devices |
| US6298148B1 (en) * | 1999-03-22 | 2001-10-02 | General Electric Company | Method of registering surfaces using curvature |
| US7076088B2 (en) * | 1999-09-03 | 2006-07-11 | Honeywell International Inc. | Near-infrared disguise detection |
| US6816605B2 (en) | 1999-10-08 | 2004-11-09 | Lumidigm, Inc. | Methods and systems for biometric identification of individuals using linear optical spectroscopy |
| US6698653B1 (en) * | 1999-10-28 | 2004-03-02 | Mel Diamond | Identification method, especially for airport security and the like |
| US7441263B1 (en) | 2000-03-23 | 2008-10-21 | Citibank, N.A. | System, method and computer program product for providing unified authentication services for online applications |
| JP2003529857A (en) * | 2000-03-30 | 2003-10-07 | ブリティッシュ・テレコミュニケーションズ・パブリック・リミテッド・カンパニー | Image processing |
| US6895111B1 (en) | 2000-05-26 | 2005-05-17 | Kidsmart, L.L.C. | Evaluating graphic image files for objectionable content |
| US6996256B2 (en) | 2000-06-08 | 2006-02-07 | Honeywell International Inc. | Detection system and method using thermal image analysis |
| US7027621B1 (en) * | 2001-03-15 | 2006-04-11 | Mikos, Ltd. | Method and apparatus for operator condition monitoring and assessment |
| US7181017B1 (en) | 2001-03-23 | 2007-02-20 | David Felsher | System and method for secure three-party communications |
| US6850147B2 (en) | 2001-04-02 | 2005-02-01 | Mikos, Ltd. | Personal biometric key |
| US6854879B2 (en) | 2001-04-19 | 2005-02-15 | Honeywell International Inc. | System and method using thermal image analysis for polygraph testing |
| US7111980B2 (en) | 2001-04-19 | 2006-09-26 | Honeywell International Inc. | System and method using thermal image analysis and slope threshold classification for polygraph testing |
| US20030048904A1 (en) * | 2001-09-07 | 2003-03-13 | Po-Tong Wang | Web-based biometric authorization apparatus |
| US7305108B2 (en) * | 2001-11-08 | 2007-12-04 | Pelco | Security identification system |
| US7136513B2 (en) * | 2001-11-08 | 2006-11-14 | Pelco | Security identification system |
| US20030154201A1 (en) * | 2002-02-13 | 2003-08-14 | Canon Kabushiki Kaisha | Data storage format for topography data |
| US7620212B1 (en) | 2002-08-13 | 2009-11-17 | Lumidigm, Inc. | Electro-optical sensor |
| DE50212936D1 (en) * | 2002-10-24 | 2008-12-04 | L 1 Identity Solutions Ag | Examination of image recordings of persons |
| US9818136B1 (en) | 2003-02-05 | 2017-11-14 | Steven M. Hoffberg | System and method for determining contingent relevance |
| JP4397212B2 (en) * | 2003-02-05 | 2010-01-13 | 富士フイルム株式会社 | Identification device |
| JP4406547B2 (en) * | 2003-03-03 | 2010-01-27 | 富士フイルム株式会社 | ID card creation device, ID card, face authentication terminal device, face authentication device and system |
| GB0306356D0 (en) * | 2003-03-20 | 2003-04-23 | Qinetiq Ltd | Biometric verification apparatus and method |
| US7668350B2 (en) | 2003-04-04 | 2010-02-23 | Lumidigm, Inc. | Comparative texture analysis of tissue for biometric spoof detection |
| US7394919B2 (en) | 2004-06-01 | 2008-07-01 | Lumidigm, Inc. | Multispectral biometric imaging |
| US7545963B2 (en) | 2003-04-04 | 2009-06-09 | Lumidigm, Inc. | Texture-biometrics sensor |
| US7751594B2 (en) | 2003-04-04 | 2010-07-06 | Lumidigm, Inc. | White-light spectral biometric sensors |
| US7627151B2 (en) | 2003-04-04 | 2009-12-01 | Lumidigm, Inc. | Systems and methods for improved biometric feature definition |
| US7460696B2 (en) | 2004-06-01 | 2008-12-02 | Lumidigm, Inc. | Multispectral imaging biometrics |
| US7539330B2 (en) | 2004-06-01 | 2009-05-26 | Lumidigm, Inc. | Multispectral liveness determination |
| JP2007524441A (en) | 2003-04-04 | 2007-08-30 | ルミディム インコーポレイテッド | Multispectral biometric sensor |
| US7347365B2 (en) | 2003-04-04 | 2008-03-25 | Lumidigm, Inc. | Combined total-internal-reflectance and tissue imaging systems and methods |
| JP4200055B2 (en) * | 2003-06-12 | 2008-12-24 | 日立オムロンターミナルソリューションズ株式会社 | Banknote transaction system |
| US7512255B2 (en) * | 2003-08-22 | 2009-03-31 | Board Of Regents, University Of Houston | Multi-modal face recognition |
| US7008386B2 (en) * | 2003-08-26 | 2006-03-07 | Acor Orthopaedic, Inc. | Foot orthotic |
| US7920725B2 (en) * | 2003-09-09 | 2011-04-05 | Fujifilm Corporation | Apparatus, method, and program for discriminating subjects |
| US20050123182A1 (en) * | 2003-12-03 | 2005-06-09 | Avision Inc. | Temperature sensor |
| US7263213B2 (en) | 2003-12-11 | 2007-08-28 | Lumidigm, Inc. | Methods and systems for estimation of personal characteristics from biometric measurements |
| WO2005064525A1 (en) * | 2003-12-30 | 2005-07-14 | Kield Martin Kieldsen | A method and apparatus for providing information relating to a body part of a person, such as for identifying the person |
| WO2005070301A1 (en) * | 2004-01-22 | 2005-08-04 | Nihondensikougaku Co., Ltd. | Male/female distinguishing means |
| JP2005259049A (en) * | 2004-03-15 | 2005-09-22 | Omron Corp | Face collation device |
| DE102004039937A1 (en) * | 2004-08-18 | 2006-02-23 | Hoffmann, André | Identification, verification and recognition method and system for human face uses visible characteristics of teeth and uses laser, camera, sensor and color |
| CA2600938A1 (en) * | 2004-03-24 | 2005-10-06 | Andre Hoffmann | Identification, verification, and recognition method and system |
| US8229185B2 (en) | 2004-06-01 | 2012-07-24 | Lumidigm, Inc. | Hygienic biometric sensors |
| US7508965B2 (en) | 2004-06-01 | 2009-03-24 | Lumidigm, Inc. | System and method for robust fingerprint acquisition |
| US20110163163A1 (en) * | 2004-06-01 | 2011-07-07 | Lumidigm, Inc. | Multispectral barcode imaging |
| US8787630B2 (en) | 2004-08-11 | 2014-07-22 | Lumidigm, Inc. | Multispectral barcode imaging |
| US7792333B2 (en) * | 2004-10-19 | 2010-09-07 | Sri International | Method and apparatus for person identification |
| US7138905B2 (en) | 2004-11-18 | 2006-11-21 | Honeywell International Inc. | Controlled environment thermal image detection system and methods regarding same |
| US7367944B2 (en) * | 2004-12-13 | 2008-05-06 | Tel Hashomer Medical Research Infrastructure And Services Ltd. | Method and system for monitoring ablation of tissues |
| US7801338B2 (en) | 2005-04-27 | 2010-09-21 | Lumidigm, Inc. | Multispectral biometric sensors |
| US20070038118A1 (en) * | 2005-08-10 | 2007-02-15 | Depue Marshall Thomas | Subcutaneous tissue imager |
| US20070071288A1 (en) * | 2005-09-29 | 2007-03-29 | Quen-Zong Wu | Facial features based human face recognition method |
| US8874477B2 (en) | 2005-10-04 | 2014-10-28 | Steven Mark Hoffberg | Multifactorial optimization system and method |
| KR20070081773A (en) * | 2006-02-13 | 2007-08-17 | 스마트 와이어레스 가부시키가이샤 | Infrared face authenticating apparatus, and portable terminal and security apparatus including the same |
| US8355545B2 (en) | 2007-04-10 | 2013-01-15 | Lumidigm, Inc. | Biometric detection using spatial, temporal, and/or spectral techniques |
| US7995808B2 (en) | 2006-07-19 | 2011-08-09 | Lumidigm, Inc. | Contactless multispectral biometric capture |
| CN103336941A (en) | 2006-07-19 | 2013-10-02 | 光谱辨识公司 | Multibiometric multispectral imager |
| US8175346B2 (en) | 2006-07-19 | 2012-05-08 | Lumidigm, Inc. | Whole-hand multispectral biometric imaging |
| US7801339B2 (en) | 2006-07-31 | 2010-09-21 | Lumidigm, Inc. | Biometrics with spatiospectral spoof detection |
| US7804984B2 (en) | 2006-07-31 | 2010-09-28 | Lumidigm, Inc. | Spatial-spectral fingerprint spoof detection |
| US8285010B2 (en) | 2007-03-21 | 2012-10-09 | Lumidigm, Inc. | Biometrics based on locally consistent features |
| WO2008130906A1 (en) * | 2007-04-17 | 2008-10-30 | Mikos, Ltd. | System and method for using three dimensional infrared imaging to provide psychological profiles of individuals |
| US7844085B2 (en) * | 2007-06-07 | 2010-11-30 | Seiko Epson Corporation | Pairwise feature learning with boosting for use in face detection |
| JP4751442B2 (en) * | 2008-12-24 | 2011-08-17 | 株式会社東芝 | Video surveillance system |
| US20110047384A1 (en) * | 2009-08-21 | 2011-02-24 | Qualcomm Incorporated | Establishing an ad hoc network using face recognition |
| DE112010003414T5 (en) | 2009-08-26 | 2012-12-06 | Lumidigm, Inc. | Biometric multiplex imaging and biometric dual imager sensor |
| CN101789078B (en) * | 2010-03-01 | 2012-07-18 | 江西财经大学 | Robust infrared face recognition technology |
| US8570149B2 (en) | 2010-03-16 | 2013-10-29 | Lumidigm, Inc. | Biometric imaging using an optical adaptive interface |
| JP5698293B2 (en) * | 2013-03-25 | 2015-04-08 | 富士フイルム株式会社 | Portable medical image display terminal and operating method thereof |
| US9322647B2 (en) * | 2013-03-28 | 2016-04-26 | International Business Machines Corporation | Determining camera height using distributions of object heights and object image heights |
| US9852324B2 (en) | 2015-12-08 | 2017-12-26 | Intel Corporation | Infrared image based facial analysis |
| CN111144331B (en) * | 2019-12-27 | 2023-06-13 | 中国计量大学 | Elbow vein image elbow median vein recognition method and elbow image acquisition device |
| CN111653012B (en) * | 2020-05-29 | 2022-06-07 | 浙江大华技术股份有限公司 | Gate control method, gate and device with storage function |
Family Cites Families (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US3611290A (en) * | 1968-06-03 | 1971-10-05 | North American Rockwell | Fingerprint minutiae reading device |
| SE365325B (en) * | 1971-11-04 | 1974-03-18 | Rothfjell R | |
| US3968475A (en) * | 1974-11-11 | 1976-07-06 | Sperry Rand Corporation | Digital processor for extracting data from a binary image |
| US4135147A (en) * | 1976-09-10 | 1979-01-16 | Rockwell International Corporation | Minutiae pattern matcher |
| DE2848325A1 (en) * | 1978-11-08 | 1980-05-22 | Philips Patentverwaltung | THERMAL IMAGING DEVICE FOR DETECTING AND DETECTING A THERMAL TARGET |
| US4445516A (en) * | 1980-05-29 | 1984-05-01 | Carl Zeiss-Stiftung | Process for the digitization and display of thermographic records |
| US4521861A (en) * | 1982-04-30 | 1985-06-04 | Texas Instruments Incorporated | Method and apparatus for enhancing radiometric imaging |
| CA1199732A (en) * | 1982-06-28 | 1986-01-21 | Koh Asai | Method and device for matching fingerprints with precise minutia pairs selected from coarse pairs |
| US4525859A (en) * | 1982-09-03 | 1985-06-25 | Bowles Romald E | Pattern recognition system |
| WO1985004088A1 (en) * | 1984-03-20 | 1985-09-26 | Joseph Rice | Method and apparatus for the identification of individuals |
| WO1989000741A1 (en) * | 1987-07-22 | 1989-01-26 | Etc. | Identification information storage and retrieval |
| US4975969A (en) * | 1987-10-22 | 1990-12-04 | Peter Tal | Method and apparatus for uniquely identifying individuals by particular physical characteristics and security system utilizing the same |
| US4868651A (en) * | 1988-05-17 | 1989-09-19 | S&S Inficon, Inc. | Digital radiography with image brightness and contrast normalization |
| GB2219905A (en) * | 1988-06-17 | 1989-12-20 | Philips Electronic Associated | Target detection system |
| US4858000A (en) * | 1988-09-14 | 1989-08-15 | A. C. Nielsen Company | Image recognition audience measurement system and method |
| US5072120A (en) * | 1989-02-09 | 1991-12-10 | Siewick Joseph T | Electromagnetic imager device |
| GB8910749D0 (en) * | 1989-05-10 | 1989-06-28 | Sherman Robin L | Methods and apparatus for obtaining information for characterising a person or animal |
| DE3919265A1 (en) * | 1989-06-13 | 1990-12-20 | Krupp Atlas Elektronik Gmbh | OPTO-ELECTRONIC OUTLOOK ASSEMBLY |
-
1991
- 1991-03-20 US US07/672,320 patent/US5163094A/en not_active Expired - Lifetime
-
1992
- 1992-03-18 JP JP4508287A patent/JP2873740B2/en not_active Expired - Fee Related
- 1992-03-18 EP EP92908784A patent/EP0533891B1/en not_active Expired - Lifetime
- 1992-03-18 CA CA002083380A patent/CA2083380C/en not_active Expired - Fee Related
- 1992-03-18 DE DE69232024T patent/DE69232024T2/en not_active Expired - Fee Related
- 1992-03-18 WO PCT/US1992/001805 patent/WO1992016910A1/en active IP Right Grant
Also Published As
| Publication number | Publication date |
|---|---|
| DE69232024T2 (en) | 2002-04-18 |
| EP0533891A4 (en) | 1993-07-07 |
| EP0533891B1 (en) | 2001-08-29 |
| US5163094A (en) | 1992-11-10 |
| WO1992016910A1 (en) | 1992-10-01 |
| EP0533891A1 (en) | 1993-03-31 |
| CA2083380A1 (en) | 1992-09-21 |
| JP2873740B2 (en) | 1999-03-24 |
| JPH06500177A (en) | 1994-01-06 |
| DE69232024D1 (en) | 2001-10-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA2083380C (en) | Method for identifying individuals from analysis of elemental shapes derived from biosensor data | |
| US6496594B1 (en) | Method and apparatus for aligning and comparing images of the face and body from different imagers | |
| US6173068B1 (en) | Method and apparatus for recognizing and classifying individuals based on minutiae | |
| Prokoski | History, current status, and future of infrared identification | |
| US8064647B2 (en) | System for iris detection tracking and recognition at a distance | |
| US5991429A (en) | Facial recognition system for security access and identification | |
| US7441123B2 (en) | Method and apparatus for characterizing and estimating the parameters of histological and physiological biometric markers for authentication | |
| EP1605822B1 (en) | Positive patient identification | |
| Prokoski et al. | Infrared identification of faces and body parts | |
| US11449590B2 (en) | Device and method for user authentication on basis of iris recognition | |
| Lin et al. | A de-identification face recognition using extracted thermal features based on deep learning | |
| JP4521086B2 (en) | Face image recognition apparatus and face image recognition method | |
| Mubeen | Biometric Authentication: Past, Present, and Future Perspectives | |
| Mahajan et al. | A Survey Paper on Iris Based Access System Using Stabilized Iris Encoding and Zernike Moments Phase Features | |
| Ibitayo et al. | Development Of Iris Based Age And Gender Detection System | |
| Aron et al. | Recognition of a Person based on the characteristics of the Iris and Retina | |
| CN111985377A (en) | Temperature measuring method and device, electronic equipment and storage medium | |
| Roshini | An Enhanced IRIS Pattern for Human Tracking in Large scale Biometric Environment | |
| Prokoski et al. | 9 INFRARED IDENTIFICATION OF | |
| Samantha | IRIS Biometric For Person Identification | |
| Shriwas et al. | INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH TECHNOLOGY | |
| Khatkale et al. | COMPARATIVE STUDY OF IRIS RECOGNITION FOR NEAR IMAGES |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| EEER | Examination request | ||
| MKLA | Lapsed |