Classifications

• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; 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
  • G06V40/13—Sensors therefor
  • G06V40/1306—Sensors therefor non-optical, e.g. ultrasonic or capacitive sensing
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; 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
  • G06V40/1365—Matching; Classification
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; 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
  • G06V40/1382—Detecting the live character of the finger, i.e. distinguishing from a fake or cadaver finger

Definitions

• the present disclosure relates to devices and methods of using multi-spectral ultrasonic imaging.
• An ultrasonic scanner may be comprised of various types of materials.
• the ultrasonic energy used in such a scanner is required to pass through most of these materials.
• the properties of the various materials through which an ultrasonic wave passes or strikes may have differing properties with regard to dispersion, diffraction, absorption and reflection such that the materials will disperse, diffract, absorb, and reflect the ultrasonic energy in different ways, and these differences may be dependent upon the wavelength of the ultrasonic energy.
• Use of a single ultrasonic frequency to image a particular object may result in limited information and detail about the object being imaged.
• tolerances may build up within the ultrasonic sensor stack that affect the signal path and may create a situation where the data collected does make use of the optimum available signal and response of the system.
• the data quality may be frequency dependent and the structural makeup of the target may present frequency dependencies.
• the basic methodology that has been applied in the prior art has been to perform a scan at a single specific frequency which maximizes the signal output as captured by a thin-film transistor (TFT) array positioned within the sensor stack.
• the single frequency may be primarily determined by the thickness and the material properties of the sensor stack and used to differentiate the fingerprint ridge and valley regions of a finger being imaged.
• the frequency determination may be made by choosing the frequency at which the sensor array output is maximized between two cases, one with the ultrasonic transmitter excitation voltage on and one with the transmitter off.
• This methodology may yield image information sets that may not match expected results in terms of fingerprint image definition in a more real-life setting. There might also be the need to tune the operational frequency throughout normal usage, which may lead to inconsistent results.
• the method may include generating an ultrasonic image information set from a plurality of pixels of the ultrasonic sensor for each of the scan frequencies.
• the image information set may include a pixel output value from each of the plurality of pixels, each pixel output value indicating an amount of energy reflected from the imaging surface.
• Each scan frequency may provide an image information set describing a plurality of pixel output signal levels associated with a fingerprint.
• Each pixel output value may indicate a signal strength that indicates an amount of ultrasonic energy reflected from a surface of a platen on which a finger is provided.
• image refers to one form of an image information set.
• the method may further comprise the step of combining the image information sets corresponding to each of the scan frequencies to generate a combined image information set.
• the combined image information set may include combined pixel output values from each of the plurality of pixels.
• Combining the image information sets may include adding pixel output values to produce a sum, dividing the sum by the number of scan frequencies to produce an average value for each of the pixels, and using the average value as the combined value.
• the term "combined" means mathematically combined.
• the method may further include using the plurality of ultrasonic image information sets to make a liveness determination and providing a liveness output signal indicating the liveness determination.
• the method may further include transforming each pixel output value to a gray-scale value and providing the gray-scale values for the plurality of pixels as the combined image information set representing the fingerprint of the finger.
• combining the image information sets includes, for each scan frequency, identifying a weighting factor, multiplying each pixel output value by the corresponding weighting factor to produce a pixel output value product, adding the pixel output value products to produce a sum, dividing the sum by the number of scan frequencies to produce an average value for each of the pixel output values, and using the average value as the combined pixel output value.
• the weighting factor may be calculated using the following equation:
• f is a lowest scan frequency
• f max is a highest scan frequency
• combining the image information sets may include creating a covariance matrix for each of the scan frequencies.
• the covariance matrix may be created from the pixel output values in the image information sets.
• the covariance matrices may be combined to provide a combined matrix having a combined value for each pixel output value.
• combining the covariance matrices comprises interpolating between entries in the covariance matrices.
• the method may include, for each scan frequency, identifying a weighting factor and multiplying each entry in the covariance matrices by the corresponding weighting factor prior to mathematically combining the covariance matrices.
• the weighting factor may be calculated using the following equation:
• w(f.) is the weighting factor for the i scan frequency
• avg. is the average value of the
• f is a lowest scan frequency
• f max is a highest scan frequency
• the method may further include the step of correlating each combined value for each of the pixels to a gray-scale value.
• the method may further include the step of providing the gray-scale values as the representation of the finger or fingerprint.
• the method may further include the step of scanning, without a finger on the imaging surface of the ultrasonic sensor, at a plurality of ultrasonic test frequencies.
• the method may further include the step of selecting one or more peak test frequencies. Each selected peak test frequency may have a reflected signal that is higher than a majority of other peak test frequencies.
• the method may further include the step of using the selected peak test frequencies as the plurality of scan frequencies. Additional scan frequencies may be identified by adding or subtracting a predetermined offset to the selected one of the peak test frequencies. In another embodiment, additional scan frequencies may be selected by identifying a range that includes the selected one of the peak test frequencies and selecting the scan frequencies to be within the identified range. In one embodiment, additional scan frequencies may be selected by identifying harmonics of the selected peak test frequency. In another embodiment, the method may further include assessing image quality of the peak test frequencies and selecting peak test frequencies having an image quality that is better than other peak test frequencies.
• One aspect of the present invention may be described as a system for generating automatically co-registered image information sets of a target object.
• the system may also be described as a system for scanning a finger.
• the system may comprise an imaging surface configured to receive a finger.
• the imaging surface may be substantially planar.
• the system may also comprise plane wave ultrasonic transmitter.
• the plane wave ultrasonic transmitter may generate one or more ultrasonic plane waves in response to a signal generator.
• the signal generator may be capable of creating electrical signals of different discrete frequencies within the ultrasonic frequency range.
• Neyman-Pearson multimodal fusion system to produce an output representation of the target object, such as an image.
• the combined image information set may include combined pixel output values from each of the plurality of pixels.
• the executable code may further comprise instructions to transform each pixel output value to a gray-scale value and provide the gray-scale values for the plurality of pixels as the combined image information set representing the fingerprint of the finger.
• the executable code may further comprise instructions to use the plurality of ultrasonic image information sets to make a liveness determination and provide a liveness output signal indicating the liveness determination.
• One aspect of the present invention may also be described as a system for scanning a finger.
• the system may comprise a means for generating one or more ultrasonic plane waves ("MFG") in response to a signal generator that is capable of creating electrical signals of different discrete frequencies within the ultrasonic frequency range.
• MFG ultrasonic plane waves
• FIG. 7 depicts a diagram showing a second configuration of a system for generating ultrasonic image information sets of an object in contact with an outer surfaced of a platen positioned on an ultrasonic sensor array;
• FIG. 11 is a flow chart illustrating a method for creating a combined image information set using two or more covariance matrices
• FIGS. 21 and 22 depict graphs illustrating various chirp sequences
• FIGS. 23 A, 23B, 24A and 24B depict graphs illustrating FFTs of chirp-coded transmitter signals
• FIG. 30 depicts sample image contours and corresponding histogram plots of a finger
• One aspect of the present invention relates generally to an ultrasonic sensor system for providing information about a target object.
• the information may be obtained from a plurality of excitation signals applied to an ultrasonic transmitter, each at a different frequency.
• a plurality of ultrasonic frequencies By using a plurality of ultrasonic frequencies, more information may be provided about a target object than may be provided by utilizing a single excitation frequency.
• Ultrasonic fingerprint sensors may function by generating and transmitting an ultrasonic wave toward a platen-type imaging surface.
• On the platen may be a target object about which information is desired.
• the desired information may be related to a fingerprint.
• Some of the ultrasonic energy reaching the platen is reflected, and this reflected energy may be detected.
• the strength of the reflected energy and the location at which it is received can be acquired.
• the acquired signals may be recorded in the form of a dataset.
• the dataset may be used to create a data stream that may be used to produce a visual image of the target object, which may be provided via a monitor or printer.
• each sensor may be slightly different in its resonant frequency and in its effects upon the ultrasonic signal passing through it. These resonant differences can show as much as a 50% change over a reasonably small change in frequency. Consequently, the same system that obtains a good output at a transmitter excitation frequency of 20 MHz may show only half of the output with a frequency of 19 MHz or 21 MHz.
• An ultrasonic image information set or dataset corresponding to the detected energy is generated and may be stored 27 for later use. That later use may include creating a data stream that causes an image of the target object to be displayed via a monitor or for fingerprint enrollment, verification and authentication.
• the process is repeated with a second frequency, and a second image information set corresponding to the detected energy is generated and may be stored for later use. This process may be repeated N times so as to create N image information sets 29.
• the plurality of image information sets may be combined to produce 28 a multi-spectral combined image information set.
• w(f.) is the weighting factor for the i scan frequency
• avg. is the average value of the scan-value data at the i scan frequency and a next lower scan frequency
• a weighting factor for the corresponding covariance matrix may be identified, and each entry in the corresponding covariance matrix may be multiplied by that weighting factor prior to mathematically combining the covariance matrices.
• the weighting factor may be calculated using the following equation:
• w(f.) is the weighting factor for the i scan frequency
• f max is a hig °hest scan freq L uency J .
• an initial scan frequency for example, the peak test frequency with the highest average amplitude or the best quality
• additional scan frequencies by adding and/or subtracting a predetermined offset to or from the initially selected scan frequency. For example, if the initially selected scan frequency has a frequency of X and the predetermined offset is Y, then a second one of the scan frequencies may be X+Y and a third one of the scan frequencies may be X-Y.
• FIG. 11 is a flow chart illustrating a method for generating a combined image information set using two or more covariance matrices.
• a first received image may be acquired 101 with a first scan frequency.
• a second received image at a second scan frequency different from the first may be acquired 103.
• interpolation may be used to estimate statistics centered around each pixel in the information set.
• an estimated image data for that pixel may be computed.
• each combined value may be correlated 140 to a gray-scale value.
• Estimated image data may be obtained by combining the results from each block of estimated image data.
• a combined representation may be obtained by combining results of the estimated image data from each set of initial image data (e.g. from various excitation frequencies).
• the gray-scale values may be provided 141 as a representation of the fingerprint.
• FIG. 14A-F Six graphs are shown in Fig. 14A-F, each indicating how an operating frequency might be selected.
• the frequency with the highest amplitude response at f r2 is selected.
• the lower left graph (Fig. 14B) frequencies with the two highest amplitude responses (f-2 and f r3 ) are selected.
• the upper middle graph (Fig. 14C) frequencies with the five highest amplitude responses (f rl through f r5 ) are selected.
• Fig. 14D frequencies with the best response quality are selected, corresponding to f r3 and f r4 .
• the upper right graph Fig.
• a first frequency "3" representing the lowest output signal with a skin-like test target applied to the sensor platen (representing a fingerprint ridge) may be determined, and a second frequency "4" representing the highest output signal with air (representing a fingerprint valley) may be determined. Note that the highest output signal with air and the lowest output signal with a skin-like test target may not always occur with the highest and lowest system peaks.
• the two determined frequencies 3 and 4 may be selected for operation. Shifts in the frequencies 3 and 4 with temperature changes may be included.
• a fourth method (Fig.
• One or more chirps may be applied in a series (e.g. repeated).
• a single information set may be acquired using a chirp sequence with multi- frequency content covering the greatest receiver frequency response.
• Multiple information sets may be acquired using one or more chirp sequences and the information sets combined.
• An ultrasonic sensor may be calibrated using these chirp sequences.
• a linear chirp signal has a frequency that changes linearly with time, for example,
• FIG. 23A and 23B depict graphs illustrating FFTs of chirp-coded transmitter signals.
• Fig. 23A shows a FFT of chirp-coded "extended chirp" transmitter signal, with a linear frequency band from 5 to 20 MHz.
• Fig. 23B shows a FFT of chirp-coded "peak-to- peak chirp” transmitter signal, with a linear frequency band from 7.5 to 12.5 MHz.
• Fig. 31 is a variability plot showing analog voltage comparison between selected ridge and valley points.
• a group of randomly selected points corresponding to ridge and valley regions of a finger is tracked for the two frequencies of operation. It can be seen that the regions representative of the ridges of a finger show the maximum change between the two operation frequencies, while the "valley" regions of the finger remain fairly unchanged.
• a standard factory-like calibration methodology can be employed by using a target material (e.g., rubber) similar in acoustic properties to a finger. Two sets of measurements may be taken, one with the target material completely covering the platen (simulating finger) and another without any target object on the platen ("air" measurement). The frequency of the tone burst signal may be swept, and the TFT response captured for both the cases (with and without the target). The difference between the two signals is then used to determine the optimal point(s) where inversion behavior is best observed which is given by the negative and positive maximum of the difference signal of air and target ("Air minus Target").
• the method may further comprise identifying a feature-value of the ridge-pixel SSHDI or FSHDI and a feature-value of the valley-pixel SSHDI or FSHDI. For each of the other information sets, the method may further comprise determining a difference between the ridge-pixel feature-value and the valley-pixel feature- value to obtain a separation value. For each of the other information sets, the method may further comprise determining whether the separation values identify a spatial location previously identified as corresponding to a live being.
• the feature-value is a signal-strength most commonly appearing in the SSHDI or FSHDI. In another embodiment, the feature-value is a median signal-strength appearing in the SSHDI or FSHDI. However, the feature-value may be a statistical energy, statistical entropy, or statistical variance of the SSHDI or FSHDI.
• the speed of sound in PVDF, parylene, and polycarbonate may be respectively as follows: m m m
• the piezoelectric layer and parylene coating on top of the piezoelectric may be observed first.
• the following equations describe a possible observation: ⁇ c pvdf c pary )
• Film(t, St, X) if(t ⁇ 6t,—X, X)(film thickness marker function)

Landscapes

• Engineering & Computer Science (AREA)
• Human Computer Interaction (AREA)
• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Multimedia (AREA)
• Theoretical Computer Science (AREA)
• Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
• Image Input (AREA)
• Ultra Sonic Daignosis Equipment (AREA)
• Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
• Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)

Priority Applications (5)

Applications Claiming Priority (4)

Publications (1)

Family

ID=52693065

Family Applications (1)

Country Status (5)

Cited By (44)

Families Citing this family (9)

Citations (7)

Family Cites Families (12)

• 2015
  • 2015-03-05 CN CN201580011084.1A patent/CN106068515B/zh active Active
  • 2015-03-05 WO PCT/US2015/019069 patent/WO2015134816A1/en not_active Ceased
  • 2015-03-05 EP EP15711007.3A patent/EP3114608A1/en not_active Withdrawn
  • 2015-03-05 JP JP2016555454A patent/JP2017514108A/ja active Pending
  • 2015-03-05 KR KR1020167024236A patent/KR20160130234A/ko not_active Abandoned

Patent Citations (7)

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events