CA2230386A1 - Automated indentification and analysis of fingerprints - Google Patents
Automated indentification and analysis of fingerprints Download PDFInfo
- Publication number
- CA2230386A1 CA2230386A1 CA002230386A CA2230386A CA2230386A1 CA 2230386 A1 CA2230386 A1 CA 2230386A1 CA 002230386 A CA002230386 A CA 002230386A CA 2230386 A CA2230386 A CA 2230386A CA 2230386 A1 CA2230386 A1 CA 2230386A1
- Authority
- CA
- Canada
- Prior art keywords
- fingerprint
- fingerprints
- ridges
- analysis
- minutiae
- 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.)
- Abandoned
Links
Classifications
-
- 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
- G06V40/12—Fingerprints or palmprints
- G06V40/1347—Preprocessing; Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Collating Specific Patterns (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
A method and apparatus for automatically identifying fingerprints through a mathematical characterization of ridge patterns.
Description
TITLE OF THE INVENTION
METHOD AND APPARATUS FOR THE AUTOMATED
IDENTIFICATION OF FINGERPRINTS.
FIELD OF THE INVENTION
The present invention relates to a method and an apparatus for automatically identifying fingerprints through a mathematical characterization of ridge patterns.
BACKGROUND OF THE INVENTION
The intricate pattern of skin ridges in fingerprints has long been considered the most reliable physical feature by which to identify an individual. Fingerprints are accepted as evidence in the courts of law, and the police maintain extensive files of fingerprints for known and accepted reasons. The governments of several U.S. states and Canadian provinces discuss, at the present, if and how fingerprinting could help to identify and authenticate the legitimate users of their social services. In the wealthy countries of the Middle East, in particular, the control of immigration and the influx of aliens as temporary workers is of the greatest concern, and the use of fingerprint technology is the only viable option answering it. Business sees a need to use fingerprints in order to grant or deny access to confidential data banks through communication networks, and to authenticate the source of decisions transmitted through such links. No other physical features characterizes an individual as reliably as fingerprints.
The use of fingerprints in any of the above applications is hampered by the fact that the characteristic pattern of skin ridges is not easily analysed by computer. Police computers dedicated to this task are among the biggest and most expensive of such machines in the public service; their capital cost is typically in excess of 4 mil. US$, to which amount has to be added the considerable expense of highly specialized personnel. Even these sophisticated machines with all their associated manpower perform at a relatively slow rate, and this is the consequence of the method of analysis developed more than 20 years ago, still the only method in use.
All current computers dedicated to the analysis of fingerprints (excepting those which use a method of direct optical comparison) identify an individual by the pattern of so-called "minutiae"
in the skin ridges. Minutiae are ridge endings, bi- and trifurcations, "eyelets" or islands, and other details. The minutiae found are typically mapped by triangulation to give a characteristic pattern to which is added, in some methods, a count of the number of ridges on the lines joining them. False minutiae (e.g. due to over-inking when the print is taken) will interfere with this analysis and can easily render it meaningless since the presence of just a few of them, or the absence of some real ones, seriously falsifies the entire network of triangles thus established.
Therefore, all analyses under the current method must be supervised and eventually corrected by the work of trained and experienced specialists who have to take the decision, among others, which of the apparent minutiae in a print are true and which are artefactual.
False minutiae can be produced in many ways: bridges between adjacent epidermal ridges are seen, for instance, if a print is over-inked or when the fingertip is dirty, or simply if an excess of pressure is applied at the time the print is taken, etc. True minutiae are lost through oversight, or by under-inking, or if the fingertip is wet. The loss of a few minutiae can be critical in the analysis of certain types of fingerprint patterns such as the simple arch and the tented arch which do not contain many minutiae to begin with.
DESCRIPTION OF PREFERRED EMBODIMENTS
It is an object of the present invention to overcome the above described problems of the presently used systems by use of a mathematical definition of the epidermal ridges forming a fingerprint pattern. This is achieved by utilizing the epidermal ridges in their entirety rather than their intercepts or endings (minutiae). This approach is just as valid in the identification of an individual as the analysis of minutiae patterns since minutiae are the necessary consequence of the convolutions of ridges. If one were to paint, e.g., the letter "U" on the surface of a viscous liquid medium and then swirl it around slightly, the lines of the letter will form junctions and possibly be interrupted. This means that the appearance of minutiae is the consequence of the distortion of the ridges in a fingerprint and it follows that tracing the ridges in a fingerprint is the topological equivalent of defining the position of its minutiae.
The present invention relates to a method of characterizing individual fingerprints which does not make use of the minutiae but instead defines its epidermal ridges in their entirety. The present method presents several obvious advantages which all derive from the fact that a much greater volume of data is used for purposes of identification.
Other objects and further scope of applicability of the present invention will become apparent from the detailed description given hereinafter. It should be understood, however, that this detailed description, while indicating preferred embodiments of the invention, is given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art.
IN THE DRAWINGS
Figure 1 is a typical fingerprint after pre-treatment for analysis;
Figure 2 is the fingerprint of figure 1, after thinning;
Figure 3 show the epidermal ridges tracked from window's diagonals; and Figures 4a and 4b show two fingerprint positions of reference points through a rotation of 90°.
DESCRIPTION OF PREFERRED EMBODIMENTS
Referring to figure 1, an image of the fingerprint to be analysed is imported into a microcomputer of the capacity and performance offered by current personal computers, through suitable and commercially available means (e.g. through scanning, by the use of a camera, etc.) Several pre-treatment ("cleaning") algorithms establish the conditions optimal for the analysis, specifically, among others:
a) contrast enhancement;
b) enhancement of ridge visibility;
c) contrast expansion;
d) levelling of image brightness.
The algorithms characterizing a fingerprint follow these procedures and perform, in sequence, the following steps:
A) Thinning: find, trace and retain the exact center line, only one pixel wide, of each of the epidermal ridges in the window picturing the fingerprint through the input device, that is, in the center of the fingertip (an area of approx. 2 X 2 cm). The purpose of this algorithm 5 is to establish the exact coordinates of the ridge intercepts (see figure 2).
B) Trace two diagonals from the corners of the square window. These lines will intersect the epidermal ridges and notably those defining the center of the pattern.
C) Tracking: follow the epidermal ridges starting from the intercepts with the diagonals until they end at any of the window's borders, or, at the confluence with another ridge. At such confluences or junctions, the tracking function continues along the ridge that follows the smaller angle with the tracked segment immediately preceding (see figure 3). The ridges intersecting the diagonals are retained, all others are discarded.
D) Definition: the traces of the ridges tracked and retained are now characterized mathematically. A proprietary algorithm is used which (a) selects from the data points defining a ridge automatically those which serve as so-called nodal points in a representation by cubic spline functions, and which, (b), combines the mathematical definition of the ridges by spline functions with a least-square fit. The coefficients of the spline functions are then assembled in the form of a matrix, in the order in which they were tracked. It is to be noted that each ridge is thus defined by its proper set of coefficients, and that it is possible to reconstruct the ridges from their mathematical description.
(E) Reference points: the maxima, minima and points of maximal slope (i.e. the inflection points) are determined for each ridge from its mathematical characterization, and tabulated. The matrix of these points identifies the fingerprint.
(F) Geometric hashing: the points retained are correlated geometrically in the form of a hashing table, i.e. the distances between each and every point to call others are measured and retained.
This approach has the particular advantage that the points are not positioned with respect to the coordinate axes of the window and that therefore the same correlations are found even if the fingerprint is oriented differently (see figure 4 which illustrates the advantage of a hashing table; the fingerprint is rotated through 90° but the relative positions of the reference points are retained).
The fact that very many fingerprints taken or found at the scene of a crime lack definition or contain different levels of noise renders the analysis as presented sometimes very difficult. In the present sequence of algorithms, several steps have been included which automatically address the problem arising from a fragmentation of the ridges and from the appearance of loops when the tracking procedure cannot decide which branch of the ridge is to be followed. In a series of tests, these algorithms proved highly successful.
To search for, and match, a fingerprint) the correlations between reference points defining a fingerprint are stored in the form of a hashing table in the computer's memory, and under the appropriate name. The hashing table of a fingerprint is simply added to those of all other fingerprints taken, for instance in police work, for a particular segment of delinquents ("juveniles"), for selected geographies, or for certain crimes or crime patterns. The computer search for a particular fingerprint will then rapidly establish how many of its reference point correlations coincide with those of all other fingerprints in the particular file, will identify the match and support it with the number of correlations ("votes") found.
Both, the matrix of spline coefficients for all epidermic ridges in a fingerprint and the hashing table can be stored and transmitted rapidly and economically. The data identifying a fingerprint can also be used for purposes of encryption: for example, selected data points derived from an individual's fingerprint can serve as a cipher to send confidential messages. The great quantity of information in a single fingerprint makes it possible to use such ciphers only once, and thus render the code virtually unbreakable. This application will be of interest in particular for businessmen who are looking for a fast, reliable and inexpensive method to transmit confidential data over unprotected phone lines.
Since the epidermal ridges are defined in a fingerprint and not only their minutiae, the total volume of data utilized by the present method is considerably greater than that of current techniques, although the volume of data retained is smaller. As a consequence, this method is much more robust, less subject to errors due to the loss of data or the appearance of unreliable or false features. Secondly, fingerprints processed by this method will only rarely require the services of trained personnel. The complete process of analysis, including the extraction of identifiers, the preparation for transmission, the search for similar or identical fingerprints on file, the comparison and the optional procedure of data encryption are performed rapidly and automatically on currently available, inexpensive PC-type microcomputers.
Although the invention has been described above with respect with one specific form, it will be evident to a person skilled in the art that it may be modified and refined in various ways. For example, the fingerprint ridges as defined by this method contain a wealth of information on their relative positions, distances, and curvatures. This information may be helpful to identify also partial (incomplete) fingerprints.
It is therefore wished to have it understood that the present invention should not be limited in scope, except by the terms of the following claims.
METHOD AND APPARATUS FOR THE AUTOMATED
IDENTIFICATION OF FINGERPRINTS.
FIELD OF THE INVENTION
The present invention relates to a method and an apparatus for automatically identifying fingerprints through a mathematical characterization of ridge patterns.
BACKGROUND OF THE INVENTION
The intricate pattern of skin ridges in fingerprints has long been considered the most reliable physical feature by which to identify an individual. Fingerprints are accepted as evidence in the courts of law, and the police maintain extensive files of fingerprints for known and accepted reasons. The governments of several U.S. states and Canadian provinces discuss, at the present, if and how fingerprinting could help to identify and authenticate the legitimate users of their social services. In the wealthy countries of the Middle East, in particular, the control of immigration and the influx of aliens as temporary workers is of the greatest concern, and the use of fingerprint technology is the only viable option answering it. Business sees a need to use fingerprints in order to grant or deny access to confidential data banks through communication networks, and to authenticate the source of decisions transmitted through such links. No other physical features characterizes an individual as reliably as fingerprints.
The use of fingerprints in any of the above applications is hampered by the fact that the characteristic pattern of skin ridges is not easily analysed by computer. Police computers dedicated to this task are among the biggest and most expensive of such machines in the public service; their capital cost is typically in excess of 4 mil. US$, to which amount has to be added the considerable expense of highly specialized personnel. Even these sophisticated machines with all their associated manpower perform at a relatively slow rate, and this is the consequence of the method of analysis developed more than 20 years ago, still the only method in use.
All current computers dedicated to the analysis of fingerprints (excepting those which use a method of direct optical comparison) identify an individual by the pattern of so-called "minutiae"
in the skin ridges. Minutiae are ridge endings, bi- and trifurcations, "eyelets" or islands, and other details. The minutiae found are typically mapped by triangulation to give a characteristic pattern to which is added, in some methods, a count of the number of ridges on the lines joining them. False minutiae (e.g. due to over-inking when the print is taken) will interfere with this analysis and can easily render it meaningless since the presence of just a few of them, or the absence of some real ones, seriously falsifies the entire network of triangles thus established.
Therefore, all analyses under the current method must be supervised and eventually corrected by the work of trained and experienced specialists who have to take the decision, among others, which of the apparent minutiae in a print are true and which are artefactual.
False minutiae can be produced in many ways: bridges between adjacent epidermal ridges are seen, for instance, if a print is over-inked or when the fingertip is dirty, or simply if an excess of pressure is applied at the time the print is taken, etc. True minutiae are lost through oversight, or by under-inking, or if the fingertip is wet. The loss of a few minutiae can be critical in the analysis of certain types of fingerprint patterns such as the simple arch and the tented arch which do not contain many minutiae to begin with.
DESCRIPTION OF PREFERRED EMBODIMENTS
It is an object of the present invention to overcome the above described problems of the presently used systems by use of a mathematical definition of the epidermal ridges forming a fingerprint pattern. This is achieved by utilizing the epidermal ridges in their entirety rather than their intercepts or endings (minutiae). This approach is just as valid in the identification of an individual as the analysis of minutiae patterns since minutiae are the necessary consequence of the convolutions of ridges. If one were to paint, e.g., the letter "U" on the surface of a viscous liquid medium and then swirl it around slightly, the lines of the letter will form junctions and possibly be interrupted. This means that the appearance of minutiae is the consequence of the distortion of the ridges in a fingerprint and it follows that tracing the ridges in a fingerprint is the topological equivalent of defining the position of its minutiae.
The present invention relates to a method of characterizing individual fingerprints which does not make use of the minutiae but instead defines its epidermal ridges in their entirety. The present method presents several obvious advantages which all derive from the fact that a much greater volume of data is used for purposes of identification.
Other objects and further scope of applicability of the present invention will become apparent from the detailed description given hereinafter. It should be understood, however, that this detailed description, while indicating preferred embodiments of the invention, is given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art.
IN THE DRAWINGS
Figure 1 is a typical fingerprint after pre-treatment for analysis;
Figure 2 is the fingerprint of figure 1, after thinning;
Figure 3 show the epidermal ridges tracked from window's diagonals; and Figures 4a and 4b show two fingerprint positions of reference points through a rotation of 90°.
DESCRIPTION OF PREFERRED EMBODIMENTS
Referring to figure 1, an image of the fingerprint to be analysed is imported into a microcomputer of the capacity and performance offered by current personal computers, through suitable and commercially available means (e.g. through scanning, by the use of a camera, etc.) Several pre-treatment ("cleaning") algorithms establish the conditions optimal for the analysis, specifically, among others:
a) contrast enhancement;
b) enhancement of ridge visibility;
c) contrast expansion;
d) levelling of image brightness.
The algorithms characterizing a fingerprint follow these procedures and perform, in sequence, the following steps:
A) Thinning: find, trace and retain the exact center line, only one pixel wide, of each of the epidermal ridges in the window picturing the fingerprint through the input device, that is, in the center of the fingertip (an area of approx. 2 X 2 cm). The purpose of this algorithm 5 is to establish the exact coordinates of the ridge intercepts (see figure 2).
B) Trace two diagonals from the corners of the square window. These lines will intersect the epidermal ridges and notably those defining the center of the pattern.
C) Tracking: follow the epidermal ridges starting from the intercepts with the diagonals until they end at any of the window's borders, or, at the confluence with another ridge. At such confluences or junctions, the tracking function continues along the ridge that follows the smaller angle with the tracked segment immediately preceding (see figure 3). The ridges intersecting the diagonals are retained, all others are discarded.
D) Definition: the traces of the ridges tracked and retained are now characterized mathematically. A proprietary algorithm is used which (a) selects from the data points defining a ridge automatically those which serve as so-called nodal points in a representation by cubic spline functions, and which, (b), combines the mathematical definition of the ridges by spline functions with a least-square fit. The coefficients of the spline functions are then assembled in the form of a matrix, in the order in which they were tracked. It is to be noted that each ridge is thus defined by its proper set of coefficients, and that it is possible to reconstruct the ridges from their mathematical description.
(E) Reference points: the maxima, minima and points of maximal slope (i.e. the inflection points) are determined for each ridge from its mathematical characterization, and tabulated. The matrix of these points identifies the fingerprint.
(F) Geometric hashing: the points retained are correlated geometrically in the form of a hashing table, i.e. the distances between each and every point to call others are measured and retained.
This approach has the particular advantage that the points are not positioned with respect to the coordinate axes of the window and that therefore the same correlations are found even if the fingerprint is oriented differently (see figure 4 which illustrates the advantage of a hashing table; the fingerprint is rotated through 90° but the relative positions of the reference points are retained).
The fact that very many fingerprints taken or found at the scene of a crime lack definition or contain different levels of noise renders the analysis as presented sometimes very difficult. In the present sequence of algorithms, several steps have been included which automatically address the problem arising from a fragmentation of the ridges and from the appearance of loops when the tracking procedure cannot decide which branch of the ridge is to be followed. In a series of tests, these algorithms proved highly successful.
To search for, and match, a fingerprint) the correlations between reference points defining a fingerprint are stored in the form of a hashing table in the computer's memory, and under the appropriate name. The hashing table of a fingerprint is simply added to those of all other fingerprints taken, for instance in police work, for a particular segment of delinquents ("juveniles"), for selected geographies, or for certain crimes or crime patterns. The computer search for a particular fingerprint will then rapidly establish how many of its reference point correlations coincide with those of all other fingerprints in the particular file, will identify the match and support it with the number of correlations ("votes") found.
Both, the matrix of spline coefficients for all epidermic ridges in a fingerprint and the hashing table can be stored and transmitted rapidly and economically. The data identifying a fingerprint can also be used for purposes of encryption: for example, selected data points derived from an individual's fingerprint can serve as a cipher to send confidential messages. The great quantity of information in a single fingerprint makes it possible to use such ciphers only once, and thus render the code virtually unbreakable. This application will be of interest in particular for businessmen who are looking for a fast, reliable and inexpensive method to transmit confidential data over unprotected phone lines.
Since the epidermal ridges are defined in a fingerprint and not only their minutiae, the total volume of data utilized by the present method is considerably greater than that of current techniques, although the volume of data retained is smaller. As a consequence, this method is much more robust, less subject to errors due to the loss of data or the appearance of unreliable or false features. Secondly, fingerprints processed by this method will only rarely require the services of trained personnel. The complete process of analysis, including the extraction of identifiers, the preparation for transmission, the search for similar or identical fingerprints on file, the comparison and the optional procedure of data encryption are performed rapidly and automatically on currently available, inexpensive PC-type microcomputers.
Although the invention has been described above with respect with one specific form, it will be evident to a person skilled in the art that it may be modified and refined in various ways. For example, the fingerprint ridges as defined by this method contain a wealth of information on their relative positions, distances, and curvatures. This information may be helpful to identify also partial (incomplete) fingerprints.
It is therefore wished to have it understood that the present invention should not be limited in scope, except by the terms of the following claims.
Claims (4)
1. A method of identifying a fingerprint by tracking automatically the epidermal ridges that define it.
2. A method of identifying a fingerprint by defining the epidermal ridges tracked through mathematical equations.
3. A method of identifying a fingerprint by expressing the epidermal ridges defined mathematically through spline functions.
4. A method of selecting and correlating certain reference points on the epidermal ridges in a fingerprint which then serve to search for, compare and match any unknown fingerprint with those retained in a file.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CA002230386A CA2230386A1 (en) | 1998-02-24 | 1998-02-24 | Automated indentification and analysis of fingerprints |
CA002262300A CA2262300A1 (en) | 1998-02-24 | 1999-02-22 | Method and apparatus for the automated identification of fingerprints |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CA002230386A CA2230386A1 (en) | 1998-02-24 | 1998-02-24 | Automated indentification and analysis of fingerprints |
Publications (1)
Publication Number | Publication Date |
---|---|
CA2230386A1 true CA2230386A1 (en) | 1999-08-24 |
Family
ID=29409331
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CA002230386A Abandoned CA2230386A1 (en) | 1998-02-24 | 1998-02-24 | Automated indentification and analysis of fingerprints |
Country Status (1)
Country | Link |
---|---|
CA (1) | CA2230386A1 (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106250857A (en) * | 2016-08-04 | 2016-12-21 | 深圳先进技术研究院 | A kind of identity recognition device and method |
CN110956468A (en) * | 2019-11-15 | 2020-04-03 | 西安电子科技大学 | Fingerprint payment system |
-
1998
- 1998-02-24 CA CA002230386A patent/CA2230386A1/en not_active Abandoned
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106250857A (en) * | 2016-08-04 | 2016-12-21 | 深圳先进技术研究院 | A kind of identity recognition device and method |
WO2018023884A1 (en) * | 2016-08-04 | 2018-02-08 | 深圳先进技术研究院 | Device and method for identity recognition |
CN110956468A (en) * | 2019-11-15 | 2020-04-03 | 西安电子科技大学 | Fingerprint payment system |
CN110956468B (en) * | 2019-11-15 | 2023-05-23 | 西安电子科技大学 | Fingerprint payment system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11934504B2 (en) | Method of host-directed illumination and system for conducting host-directed illumination | |
US10437970B2 (en) | User authentication and data encryption | |
US20160269411A1 (en) | System and Method for Anonymous Biometric Access Control | |
TW201546645A (en) | Method and system for verifying identities | |
Padma et al. | A survey on biometric based authentication in cloud computing | |
EP3655874B1 (en) | Method and electronic device for authenticating a user | |
CN105337739B (en) | Safe login method, device, server and terminal | |
CN107196971A (en) | Information processing method, device, electronic equipment and server | |
Lone et al. | A novel OTP based tripartite authentication scheme | |
CA2230386A1 (en) | Automated indentification and analysis of fingerprints | |
Kiran et al. | Biometric authentication: a holistic review | |
McDowall | Biometrics: The password you'll never forget | |
CA2262300A1 (en) | Method and apparatus for the automated identification of fingerprints | |
Chattopadhyay et al. | Information Assurance and Security Issues in Telemedicine—Future Directions | |
CN110012034B (en) | Biological characteristic authentication method with privacy protection function in cloud environment | |
CN116644403A (en) | Security authentication method based on security computer | |
Mather | From Scotland Yard to touchless authentication–fingerprinting makes its mark | |
CN111475790A (en) | Authentication management method and electronic equipment | |
JP2003345758A (en) | Individual authentication device and method | |
Dennis | The Wireless Ubiquitous Surveillance Testbed | |
Ford et al. | The wireless ubiquitous surveillance testbed |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
FZDE | Discontinued |