The present invention relates generally to imaging, and more specifically, to a fingerprint imaging method and system.
Biometrics is the science of using unique personal and physical characteristics as a way to verify the claimed identity of a person. One of the oldest biometric sciences involves the use of fingerprints to identify a person. Since fingerprints are formed during fetal development, and are considered to remain unchanged throughout one's life, except for the rare cases where the fingerprint may change due to diseases or scars.
It is widely accepted that no fingers have identical prints, even from the same person or from identical twins. The systematic classification of fingerprints began in the 1800s and has been developed since that time through the extensive use of fingerprints by law enforcement and forensic societies.
In current usage, a fingerprint comparison is based on minutia (i.e., individual characteristics within the fingerprint pattern). For example, these minutia can include ridge endings and bifurcations in the fingerprint. The relative position of minutia is used for comparison, and according to empirical studies, two individuals typically will not have eight or more common minutia. A typical scan of a fingerprint contains 30-40 minutia.
Since fingerprints have a large amount of data, fingerprint biometrics have a high inherent accuracy that enables low rates of false acceptance (when a biometrics system wrongly verifies the identity of a person) and false rejection (when a biometrics system is unable to verify the legitimate claimed identity of a person).
In recent years, there has been much interest in developing fingerprint biometrics systems. For example, there are several commercially available fingerprint systems. One system is available from Veridicom of Santa Clara, Calif. A second system is available from IDEX AS of Heggedal, Norway.
An important aspect of any fingerprint biometrics system is the accurate capture of the fingerprint image. To ensure accurate capture, one approach taken by companies such as the Veridicom, is to have an array of sensors that covers the entire area of the fingerprint. One disadvantage of these full-array type of finger print imagers is that these imagers require a large sensor array to perform the image capture. This prior art fingerprint imager typically employs a sensor array whose area is commensurate with the area of the entire fingerprint. The large sensor array requires a sensor for each point of the fingerprint. For example, a standard sensor array has 400,000 sensors arranged as 500 pixels by 800 pixels so that the entire image of the fingerprint can be captured at one time.
It is well known that the cost of an integrated circuit is a function of the size or area of the integrated circuit. It would be desirable for there to be a mechanism to image a fingerprint by using an array of sensors that has fewer sensors than the number of sensors needed to cover the entire area of the finger print. By reducing the number of sensors needed to accurately capture the fingerprint image, the area of the integrated circuit can be decreased, thereby leading to cost savings. Unfortunately, the prior art approaches fail to provide such a mechanism.
A second approach followed by companies, such as IDEX AS of Heggedal, Norway, is to use a scan approach where a user is required to move his finger across an array or sensors that is substantially less than the total height of the finger. This approach is described in PCT patent application, International Publication Number WO 99/43258, entitled “Fingerprint Sensor.” Unfortunately, this scan-type imager has several drawbacks. First, the IDEX approach uses a curved sensor and sliding assemblies (see
PCT patent application, International Publication Number WO 98/58342, entitled “Method and Apparatus for Measuring Structures in a Fingerprint,” further describes a system in which a scan-type sensor can be utilized. This system employs a sensor array with two or more sensor lines in order to measure structures in the fingerprint smaller than the spacing of the sensors. Since this system uses two or more lines of sensors to image a single line of the fingerprint, the system must address the problem of combining these measurements of different lines into a single signal representing the single line in the fingerprint.
In order to combine these signals into a single signal, the patent application describes different correlation techniques that can be employed to combine signals measured from different lines of sensors. The system uses an adjustment scheme that corrects for the time delay between the signals from the sensors in different lines. Unfortunately, these techniques may increase the cost and complexity of the system. Consequently, it would be desirable to have a system that employs a single row of sensors that has comparable resolution and at the same time does not require complex signal correlation techniques.
Another problem faced by scan-type systems that are not encountered with full-array type sensor systems is that the movement of the finger needs to be determined. For example, to increase the accuracy of the captured image, the movement of the finger with respect to the sensor needs to be determined. Prior art systems perform this determination by using software executing on a host computer that is remote from the sensor. Unfortunately, the host computer solution may not adequately be able to handle the rate of data transfer between the sensor chip and the host computer unless a high-speed bus, which may be prohibitively expensive to implement, is utilized. Consequently, it is desirable for a system to have a single chip solution that has a sensor array and a mechanism for determining the movement of the finger.
Based on the foregoing, there remains a need for an apparatus and method for capturing an image of a finger print that employs a single sensor chip having an integrated navigation engine for determining the change in position of the finger with respect to time.
A method and system of imaging an object. First, movement information of the object is determined by using a navigation sensor array and a navigation engine. Based on the movement information, an imaging sensor strobe signal is selectively asserted to control an imaging sensor array. In response to an asserted imaging sensor strobe signal, the imaging sensor array captures sub-images of the object. The imaging sensor array has a plurality of pixels for imaging a portion of the object at one time. The plurality of sub-images is captured successively by the imaging sensor array as the object moves with respect to the imaging sensor array. Then, a composite image of the object is generated based on the captured sub-images.
The present invention is illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which like reference numerals refer to similar elements.
a) is a block diagram of the fingerprint imager configured in accordance with one embodiment of the present invention for an imaging application.
a) and 4(b) illustrate exemplary captured and corrected fingerprint images, respectively.
a) and 5(b) illustrate exemplary captured and corrected fingerprint images, respectively.
a), 8(b), 8(c) and 8(d) illustrate how fingerprint imager of the present invention determines whether a finger has moved a predetermined distance in a first direction or a predetermined distance in a second direction in accordance with a one embodiment of the present invention.
A method and apparatus for imaging a fingerprint are described. In the following description, for the purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the present invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced without these specific details. In other instances, well-known structures and devices are shown in block diagram form in order to avoid unnecessarily obscuring the present invention.
One novel aspect of the present invention, is the use of an imaging array that captures sub-images or portions of the object to be imaged (e.g., a fingerprint) at one time and a mechanism for generating a composite image of the object (e.g., a fingerprint) by using the captured sub-images.
Referring to
It is noted that the ISA 204 can be configured in alternative arrangements. For example, instead of a single row or sensors, a predetermined number (M) of rows of sensors can be employed. Furthermore, if the fingerprint imager is configured for capturing fingerprints when a finger is swiped in a generally left-to-right motion, one or more columns of sensors can be utilized. It is noted that the number of sensors and the orientation of those sensors can be adjusted to suit the requirements of a particular application. These alternative arrangements are described in greater detail with reference to
Referring to
In operation, a user positions the user's finger on an imaging surface and pulls or rolls the finger across the imaging surface in a direction of movement. The imaging surface can be a physical surface, such as a glass plane, or an optical imaging plane. It is noted that the teachings of the present invention can be applied equally to any type of imaging sensor, such as the optical type sensors, resistive-type sensors, and capacitive type sensors.
As described in greater detail hereinafter, the actual direction of movement can be represented by a vector of movement that can be decomposed into a first component in a first direction (e.g., a change in the x direction) and a second component in a second direction (e.g., a change in the y direction).
It is noted that in this embodiment, the finger is moved in a generally downward vertical direction. Each sub-image or stripe (1 to N) is captured successively as the finger moves in the generally downward vertical direction as indicated by the arrow. The sub-images are subsequently connected to form a composite image of the fingerprint.
Fingerprint Imager 400
a) is a block diagram of the fingerprint imager 400 configured in accordance with one embodiment of the present invention. The fingerprint imager 400 includes a sensor integrated circuit 404 (hereinafter referred to a sensor chip) that is described in greater detail hereinafter and a host device 460 (e.g., a host personal computer (PC)) coupled thereto. As described in greater detail hereinafter, the sensor chip 404 can be configured to implement different applications by using different software applications executing on the host PC 460.
One aspect of the present invention is the provision of a sensor chip 404 having an integrated navigation engine 420 for determining the change of position (e.g., a change in the x direction and a change in the y direction) of the finger. By integrating the navigation engine 420 into the sensor chip and providing a single chip solution, the present invention increases the processing speed, as compared to using software on the host for similar calculations, decreases costs, and obviates the need for a high speed bus or connection between the host 460 and the sensor chip 404.
The sensor chip 404 includes an image sensor array (ISA) 204 for imaging a portion of the fingerprint at one time and a navigation sensor array (NSA) 410 for capturing a navigation image of the fingerprint in response to an asserted navigation sensor array (NSA) strobe signal 414. The rate of the NSA strobe signal 414 can be predetermined and is preferably in the range of 1500 to 2000 frames per second. It is noted that the rate of the NSA strobe signal 414 can be varied to suit a particular application.
In one embodiment, the navigation sensor array (NSA) 410 has a plurality of pixels arranged in a square area (e.g., an array of sensors arranged in a square area having about 20 sensors by about 20 sensors). It is noted that the pixels in the NSA 410 can be arranged in an area with other geometric shapes, such as a generally rectangular shape, a generally circular shape or generally elliptical shape. The number of pixels in the NSA 410 and the specific arrangement of the pixels in the NSA 410 can be varied to suit a particular application.
As noted previously, the fingerprint imager 400 also includes a navigation engine 420 for selectively asserting the navigation sensor array (NSA) strobe signal, receiving the navigation image from the NSA 410, and responsive thereto for generating movement information. The navigation engine 420 determines when an object to be imaged (e.g., a fingerprint) has moved a predetermined unit distance with respect to the NSA 410. The unit distance can vary across applications and be expressed in terms of resolution, such as, one pixel, ½ of a pixel, 1/10 of a pixel, etc. The navigation engine 420 causes the rate at which the ISA 204 captures images of the object to depend on the speed of movement of the object to be imaged.
U.S. Pat. No. 5,729,008, entitled “Method and Device for Tracking Relative Movement by Correlating Signals from an Array of Photoelements,” which is incorporated by reference herein, describes in greater detail the operation and an exemplary implementation of the navigation engine 420 and the NSA 410.
The prior art solutions are limited to a constant predetermined rate of image capture, which as described earlier, causes a greater likelihood of errors in the resulting image. In contrast, the present invention captures images of the object at a varying rate based on the velocity of the moving object (e.g., the finger) to provide evenly spaced sub-images. These sub-images can then be assembled into the composite image by using software that is well known by those of ordinary skill in the art.
The fingerprint imager 400 also includes the imaging sensor array (ISA) 204 for capturing a portion of an image of the fingerprint in response to an asserted imaging sensor array (ISA) strobe signal 444. In one embodiment, the imaging sensor array (ISA) 440 includes a 1 by about 350 sensors. The number of pixels utilized can be varied to suit a particular application and to achieve a desired resolution. For example, if 500 dots per inch (dpi) is desired, and the object to image is a finger, which has a width generally of about 0.75 inches, about 350 to about 400 are needed to provide such a resolution. It is noted that the size of the sensors utilized in the ISA 204 can be the same pixel size or of a different pixel size than those sensors utilized in the NSA 410.
When the NSA 410 and ISA 204 are configured in accordance to the first embodiment of the present invention, it is preferable that the sensors employed by the NSA 410 and the ISA 204 have the same dimensions or pitch density. For example, all the sensors can have dimensions of 50 microns by 50 microns. This first embodiment that has the sensor-sharing configuration has the advantage of further reducing the cost of manufacturing the fingerprint imager of the present invention by reducing the number of sensors needed to implement the imager.
When the NSA 410 and ISA 204 are configured in accordance to the second embodiment of the present invention, the dimensions of the sensors in NSA 410 and ISA 204 can be of different dimensions or pitch densities. In particular, it may be preferable for certain applications to have the sensors of the NSA 410 with dimensions small than the sensors of the ISA 204 in order for the NSA 410 to have better resolution than the ISA 204. For example, the sensors for the ISA 204 can have dimensions of 50 microns by 50 microns, whereas the sensors for the NSA 410 can have dimensions of 20 microns by 20 microns.
The fingerprint imager 400 also includes an ISA controller 430 for receiving the movement information from the navigation engine 420, and responsive thereto for selectively asserting the ISA strobe signal. As noted previously, an asserted ISA strobe signal causes the ISA 204 to image a portion of the fingerprint (i.e., capture a portion of the fingerprint image).
Authentication Application
The host 460 can include a composite image generation program 464 that receives the scans or sub-images, and the displacement of each sub-image with respect to a previous sub-image, and based thereon generates a composite image of the fingerprint that is corrected for skew due to uneven movement or “roll” of the finger as it moves with respect to the ISA 204. The host 460 can also include an authentication application 470 that extracts minutia information from the fingerprint image, compares the minutia information with previously-stored minutia information, and grants access to a resource (e.g., a computer system or a file) if there is a match between the minutia of the scanned fingerprint and one of the previously-stored minutia.
Cursor Control Application
b) illustrates an alternative embodiment where the imager of the present invention is utilized to implement a cursor control application. In this application, the ISA 204 and ISA strobe generator 430 are not needed. The host 460 can include a cursor control application 480 that ignores the scans or sub-images generated by the ISA 204 and uses only the movement information generated by the navigation engine 420 to control the cursor. In this application, the NSA 410 serves to function as a mini-touch pad, and the navigation engine 420 is utilized to determine a desired movement of a cursor.
If an object to be imaged is not detected, step 504 is repeated until an object is detected. If an object to be imaged is detected, in step 508 a navigation image is captured of the object. Based on the navigation image and previously captured navigation images, movement information of the object is determined. For example, as described in greater detail hereinafter, the change in x-position and y-position of the object can be determined by the navigation images.
In step 514, a determination is made whether the movement or change of position is in a first predetermined relationship with predefined criteria. The change in position can be determine with respect to a current image provided by the NSA 410 and a previously captured image (e.g., an immediately preceding image). If the movement or change of position is in a first predetermined relationship with predefined criteria, then in step 518 an imaging sensor array strobe signal is asserted, and a sub-image of the object is captured (e.g., one of the stripes shown in
In one embodiment, step 514 determines whether the change in x position is greater than or equal to a unit distance in a first direction (e.g., predetermined delta x) or whether the change in y position is greater than or equal to a unit distance in a second direction (e.g., predetermined delta y). If so, then the ISA strobe signal is asserted to capture a sub-image of the object. Otherwise, no sub-image is captured.
For example, the ISA strobe generator 430 can use the value of movement in a first direction and the value of movement in a second direction to determine if the finger has moved a predetermined unit distance in either direction. If it is determined that the finger has moved a predetermined distance in either direction, then the ISA strobe generator 430 asserts an ISA strobe 444 that selectively controls the ISA 204 to image a portion of the fingerprint. If it is determined that the finger has not moved a predetermined distance in either direction, then the ISA strobe generator 430 does not assert the ISA strobe signal 444 and the ISA 204 does not image a portion of the fingerprint.
If the movement or change of position is not in a first predetermined relationship with predefined criteria, then processing continues at step 508. Steps 508 and 514 are repeated until the movement or change of position is in the first predetermined relationship with the predefined criteria. In step 524, the sub-images captured in step 518 are provided to a host device, such as a personal computer (PC).
In step 650, minutia features are extracted from the object, when the object is a fingerprint. In step 660, the minutia features are compared with previously stored minutia features. In step 670, the comparison of step 660 is utilized to determine whether to grant access privileges to a device, file, network etc. or to otherwise authenticate a user's identity.
a) and 4(b) illustrate exemplary captured fingerprint images and corrected fingerprint images, respectively.
The correction software uses the movement information (e.g., x and y information) provided by the navigation engine 420 to correct for the skew in the image and generate a composite image where the sub-images are generally stacked evenly one on top the other. For example, the x offset and the y offset of each scan or sub-image with respect to a previous scan or sub-image is utilized to correct for any skew introduced by the movement of the finger. In this manner, each scan or sub-image is stacked upon the previous scan or sub-image to generate images that are generally upright and have a rectangular appearance instead of a trapezoidal appearance.
a) and 5(b) illustrate exemplary captured fingerprint images and corrected fingerprint images, respectively.
a), 8(b), 8(c) and 8(d) illustrate how fingerprint imager of the present invention determines whether a finger has moved a predetermined distance in a first direction or a predetermined distance in a second direction in accordance with a one embodiment of the present invention. In one embodiment, step 514 includes the following sub-steps. First, a first navigation image corresponding to a first time is captured. Second, a second navigation image corresponding to a second time is captured. Third, the first and second navigation images are compared to derive movement information. The movement information can be represented by a vector that has both a magnitude and direction.
For example, referring to
a illustrates the case where the x component is less than the predetermined x unit value, but the y component is greater than or equal to the predetermined y unit value.
In the foregoing specification, the invention has been described with reference to specific embodiments thereof. It will, however, be evident that various modifications and changes may be made thereto without departing from the broader scope of the invention. The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.
Number | Name | Date | Kind |
---|---|---|---|
4047154 | Vitols et al. | Sep 1977 | A |
4747147 | Sparrow | May 1988 | A |
4784484 | Jensen | Nov 1988 | A |
5195145 | Backus et al. | Mar 1993 | A |
5644139 | Allen et al. | Jul 1997 | A |
5729008 | Blalock et al. | Mar 1998 | A |
5838306 | O'Connor et al. | Nov 1998 | A |
5854482 | Bidiville et al. | Dec 1998 | A |
5869822 | Meadows et al. | Feb 1999 | A |
5920384 | Borza | Jul 1999 | A |
6002815 | Immega et al. | Dec 1999 | A |
6052475 | Upton | Apr 2000 | A |
6097035 | Belongie et al. | Aug 2000 | A |
6207945 | Bohn et al. | Mar 2001 | B1 |
6249360 | Pollard et al. | Jun 2001 | B1 |
6256022 | Manaresi et al. | Jul 2001 | B1 |
6259108 | Antonelli et al. | Jul 2001 | B1 |
6282303 | Brownlee | Aug 2001 | B1 |
6289114 | Mainguet | Sep 2001 | B1 |
6317508 | Kramer et al. | Nov 2001 | B1 |
6333989 | Borza | Dec 2001 | B1 |
6337918 | Holehan | Jan 2002 | B1 |
6360004 | Akizuki | Mar 2002 | B1 |
6408087 | Kramer | Jun 2002 | B1 |
6643389 | Raynal et al. | Nov 2003 | B1 |
6677929 | Gordon et al. | Jan 2004 | B2 |
6886104 | McClurg et al. | Apr 2005 | B1 |
7043061 | Hamid et al. | May 2006 | B2 |
7099496 | Benkley, III | Aug 2006 | B2 |
7116805 | Machida | Oct 2006 | B2 |
7203347 | Hamid | Apr 2007 | B2 |
Number | Date | Country |
---|---|---|
0813164 | Dec 1997 | EP |
0923018 | Jun 1999 | EP |
0923018 | Jun 1999 | EP |
WO 8606266 | Nov 1986 | WO |
WO 9858342 | Dec 1998 | WO |
WO 9935964 | Jul 1999 | WO |
WO 9943258 | Sep 1999 | WO |