WO2019104932A1 - 一种指纹识别方法及指纹采集设备 - Google Patents

一种指纹识别方法及指纹采集设备 Download PDF

Info

Publication number
WO2019104932A1
WO2019104932A1 PCT/CN2018/083528 CN2018083528W WO2019104932A1 WO 2019104932 A1 WO2019104932 A1 WO 2019104932A1 CN 2018083528 W CN2018083528 W CN 2018083528W WO 2019104932 A1 WO2019104932 A1 WO 2019104932A1
Authority
WO
WIPO (PCT)
Prior art keywords
fingerprint
fingerprint image
sweat
sweat hole
image
Prior art date
Application number
PCT/CN2018/083528
Other languages
English (en)
French (fr)
Inventor
任志浩
王升国
刘文杰
张政
Original Assignee
杭州海康威视数字技术股份有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 杭州海康威视数字技术股份有限公司 filed Critical 杭州海康威视数字技术股份有限公司
Publication of WO2019104932A1 publication Critical patent/WO2019104932A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition

Definitions

  • the present application relates to the field of image processing technologies, and in particular, to a fingerprint identification method and a fingerprint collection device.
  • Fingerprint recognition can be applied in many fields, such as access control, device unlocking, and payment.
  • the fingerprint collection device collects the fingerprint image of the user
  • the feature of the detail point in the fingerprint image such as the bifurcation point and the endpoint of the fingerprint ridge line in the fingerprint image, may be extracted, and the fingerprint is identified according to the extracted feature point feature.
  • the minutiae information in the user's fingerprint is easily leaked.
  • the user's fingerprint is forged using the minutiae feature in the fingerprint.
  • This forged fingerprint is very low cost, but it can produce fairly clear fingerprint detail features. Therefore, the above fingerprint identification method has low accuracy when identifying authentic fingerprints, and there is a security risk.
  • the purpose of the embodiment of the present application is to provide a fingerprint identification method and a fingerprint collection device to improve the accuracy and security of fingerprint recognition.
  • the specific technical solution is as follows.
  • an embodiment of the present application provides a fingerprint identification method.
  • the method includes:
  • Identifying a fingerprint in the fingerprint image to be identified according to the sweat hole feature Identifying a fingerprint in the fingerprint image to be identified according to the sweat hole feature.
  • the step of performing interpolation on the fingerprint image to be identified includes:
  • the fingerprint image to be identified is interpolated by using one of a bilinear interpolation algorithm and a bicubic interpolation algorithm.
  • the step of detecting the sweat hole area in the first fingerprint image according to the preset sweat hole pixel feature includes:
  • a connected domain having a number of pixel points within a preset number range is determined as a sweat hole region in the first fingerprint image.
  • the step of identifying the fingerprint in the fingerprint image to be identified according to the characteristics of the sweat hole includes:
  • the sweat hole feature comprises: at least one of a number of sweat holes and a distribution density of sweat holes.
  • the step of identifying the fingerprint in the fingerprint image to be identified according to the characteristics of the sweat hole includes:
  • the step of determining a sweat hole feature of the first fingerprint image according to the detected sweat hole region includes:
  • the step of determining a sweat hole feature of the first fingerprint image according to the detected sweat hole region includes:
  • the step of determining a target sweat hole region by using a sweat hole region centered on each detail point and having a radius of a preset distance includes:
  • the minutiae point For each minutiae point, if the number of all sweat holes in the range centered on the minutiae and the radius of the preset distance is greater than a preset number threshold, then the minutiae will be centered It is determined that all the sweat hole regions within the range of the radius are determined as the target sweat hole regions.
  • the step of detecting the sweat hole region in the first fingerprint image according to the preset sweat hole pixel feature comprises: binarizing the first fingerprint image to obtain a binarized fingerprint image, and detecting Determining the first fingerprint image when each connected domain in the binarized fingerprint image determines a connected domain having a number of pixel points within a preset number range as a sweat hole region in the first fingerprint image
  • the steps of the detail points on the fingerprint ridgeline include:
  • a detail point on the fingerprint ridge line is determined from the refined image as a detail point on the fingerprint ridge line in the first fingerprint image.
  • an embodiment of the present application provides a fingerprint collection device.
  • the device includes: a sensor, a processor, and a memory;
  • the sensor is configured to collect a fingerprint image to be identified
  • the processor is configured to acquire a fingerprint image to be recognized by the sensor, interpolate the fingerprint image to be identified, obtain a first fingerprint image, and detect the first fingerprint image according to a preset sweat pixel feature. Identifying a sweat hole feature of the first fingerprint image according to the detected sweat hole region, and identifying a fingerprint in the fingerprint image to be recognized according to the sweat hole feature; wherein, the The resolution of a fingerprint image is greater than the resolution of the fingerprint image to be identified.
  • the processor is specifically configured to perform interpolation on the fingerprint image to be identified by using one of a bilinear interpolation algorithm and a bicubic interpolation algorithm.
  • the processor is specifically configured to:
  • the connected domain is determined to be a sweat hole region in the first fingerprint image.
  • the processor is specifically configured to: determine, according to the sweat hole feature and the preset first sweat hole feature, whether the fingerprint in the fingerprint image to be identified is a living fingerprint.
  • the sweat hole feature comprises: at least one of a number of sweat holes and a distribution density of sweat holes.
  • the processor is specifically configured to: determine, according to the sweat hole feature, a pre-stored second sweat feature and an object, an object corresponding to the sweat hole feature, as the fingerprint image to be recognized The object to which the fingerprint belongs.
  • the processor is specifically configured to determine a sweat hole feature of the first fingerprint image according to at least one of the following: a center point coordinate of the sweat hole region, a size of the sweat hole region, and a sweat hole region.
  • the direction field of the fingerprint ridge is specifically configured to determine a sweat hole feature of the first fingerprint image according to at least one of the following: a center point coordinate of the sweat hole region, a size of the sweat hole region, and a sweat hole region.
  • the direction field of the fingerprint ridge is specifically configured to determine a sweat hole feature of the first fingerprint image according to at least one of the following: a center point coordinate of the sweat hole region, a size of the sweat hole region, and a sweat hole region. The direction field of the fingerprint ridge.
  • the processor is specifically configured to:
  • Determining each detail point on the fingerprint ridge line in the first fingerprint image determining a target sweat hole area and extracting the target sweat hole by using a sweat hole area centered on each detail point and having a radius of a preset distance A feature of the region as a sweat feature of the first fingerprint image.
  • the processor is specifically configured to:
  • the minutiae point For each minutiae point, if the number of all sweat holes in the range centered on the minutiae and the radius of the preset distance is greater than a preset number threshold, then the minutiae will be centered It is determined that all the sweat hole regions within the range of the radius are determined as the target sweat hole regions.
  • the processor is specifically configured to:
  • Detecting the sweat hole region in the first fingerprint image according to the preset sweat pixel feature includes: binarizing the first fingerprint image to obtain a binarized fingerprint image, and detecting the binary fingerprint Each of the connected domains in the image, when the connected domain of the number of pixels in the preset number of ranges is determined as the sweaty region in the first fingerprint image, according to the sweat hole region in the first fingerprint image, Filling the binarized fingerprint image to obtain a filled image; refining the filled image to obtain a refined image; determining a detail point on the fingerprint ridge line from the refined image as the first fingerprint image The detail point on the fingerprint ridgeline.
  • an embodiment of the present application provides a computer readable storage medium.
  • the computer readable storage medium stores a computer program, and when the computer program is executed by the processor, the fingerprint identification method provided by the embodiment of the present application is implemented.
  • the method includes:
  • Identifying a fingerprint in the fingerprint image to be identified according to the sweat hole feature Identifying a fingerprint in the fingerprint image to be identified according to the sweat hole feature.
  • the fingerprint identification method and the fingerprint collection device may perform interpolation on the fingerprint image to obtain a first fingerprint image having a resolution greater than that of the fingerprint image to be recognized, and according to the preset characteristics of the sweat pixel, the detection a sweat hole region in a fingerprint image, and determining a sweat hole feature of the first fingerprint image according to the sweat hole region, and identifying the fingerprint in the fingerprint image according to the sweat hole feature.
  • the fingerprint image to be recognized may be interpolated to obtain a fingerprint image with a higher resolution, so that the sweat hole region can be detected more accurately from the fingerprint image; the sweat hole of the fingerprint image can be determined according to the detected sweat hole region.
  • the feature is to perform fingerprint recognition on the fingerprint image to be recognized according to the characteristics of the sweat hole. Since the sweat hole feature on the fingerprint ridge line is more complicated and more difficult to be copied than the detail point, it can improve the accuracy of fingerprint recognition when applied to fingerprint recognition. Of course, implementing any of the products or methods of the present application does not necessarily require that all of the advantages described above be achieved at the same time.
  • FIG. 1 is a schematic flowchart of a fingerprint identification method according to an embodiment of the present application.
  • FIG. 2a is a schematic diagram of a fingerprint image to be identified according to an embodiment of the present application.
  • 2b is a schematic diagram of an interpolated fingerprint image provided by an embodiment of the present application.
  • 2c is a schematic diagram of a binarized image provided by an embodiment of the present application.
  • FIG. 2 is a schematic diagram of eight adjacent pixel points around a pixel point according to an embodiment of the present application
  • FIG. 3 is another schematic flowchart of a fingerprint identification method according to an embodiment of the present disclosure
  • FIG. 4 is still another schematic flowchart of a fingerprint identification method according to an embodiment of the present application.
  • FIG. 5a is a schematic diagram of screening a sweat hole region according to an embodiment of the present application.
  • FIG. 5b is a schematic diagram of a refined image provided by an embodiment of the present application.
  • FIG. 5c and FIG. 5d are schematic diagrams of bifurcation points and endpoints in the refined image provided by the embodiment of the present application.
  • FIG. 6 is a schematic structural diagram of a fingerprint collection device according to an embodiment of the present disclosure.
  • the fingerprint collection device can perform fingerprint identification on the fingerprint image.
  • the fingerprint image used by the fingerprint collection device for fingerprint recognition is a low-resolution fingerprint image, for example, a fingerprint image with a resolution of 500 to 1000 dpi.
  • the fingerprint recognition may be performed according to the feature of the detail point in the fingerprint image.
  • the feature of the detail point is easily forged by a malicious person, thereby producing a fake fingerprint, which results in the accuracy and security of the fingerprint collection device for fingerprint recognition.
  • the embodiment of the present application provides a fingerprint identification method and a fingerprint collection device.
  • the present application will be described in detail below through specific embodiments.
  • FIG. 1 is a schematic flowchart of a fingerprint identification method according to an embodiment of the present application.
  • the method is applied to an electronic device having a fingerprint recognition function, such as a fingerprint collection device.
  • the method includes:
  • Step S101 Acquire a fingerprint image to be identified.
  • the electronic device may have a fingerprint collection function or may not have a fingerprint collection function.
  • the electronic device when the electronic device has the fingerprint collection function, when the fingerprint image to be recognized is acquired, the acquired fingerprint image to be recognized may be directly acquired.
  • the fingerprint image to be recognized can be obtained from other devices.
  • the fingerprint image to be identified is a low resolution fingerprint image.
  • the fingerprint image to be recognized in this embodiment can be understood as including a fingerprint sweat hole, but a fingerprint image of a sweat hole region sufficient for fingerprint recognition cannot be directly detected from the fingerprint image to be recognized.
  • the fingerprint sweat hole is the sweat gland opening on the skin surface, generally located at the top of the undulating ridge on the skin surface, and belongs to the third layer feature of the fingerprint.
  • the wavy ridges on the surface of the skin are fingerprint ridges in the fingerprint image.
  • the fingerprint image to be identified may be a fingerprint image whose resolution is within a preset resolution range.
  • a fingerprint image whose resolution is within the preset resolution range is a low resolution image.
  • the preset resolution range can be 500 to 1000 dpi or other range.
  • the fingerprint image to be identified may be a fingerprint image with a resolution of 500 to 1000 dpi.
  • Fingerprint images below 500 dpi are generally considered to have almost no sweat hole information, and fingerprint images higher than 1000 dpi need to be collected by a fingerprint collection device with a higher configuration.
  • the upper limit of the resolution of the low-resolution fingerprint image of 500 dpi and the upper limit of the resolution of 1000 dpi are only examples.
  • the lower limit of the resolution may also be 499 dpi, 498 dpi, and the upper limit of the resolution may also be 1001 dpi. , 1002dpi equivalent.
  • the above limitation of the resolution and the upper limit of the resolution are not specifically limited in the present application.
  • the embodiment can be applied to a common access fingerprint device.
  • the resolution of the fingerprint image collected by such an access fingerprint device is usually between 500 dpi and 1000 dpi.
  • FIG. 2a is a fingerprint image acquired by a 500 dpi optical fingerprint collection device, and the white point on the fingerprint ridge line in the figure is a sweat hole.
  • the fingerprint image carries the fingerprint sweat hole, but the pixel of the sweat hole in the fingerprint image is not obvious and cannot be directly used for detection or comparison. In this case, step S102 can be performed.
  • Step S102 Interpolating the fingerprint image to be recognized to obtain a first fingerprint image.
  • the resolution of the first fingerprint image is greater than the resolution of the fingerprint image to be identified.
  • the resolution of the fingerprint image to be identified is 500 to 1000 dpi
  • the resolution of the first fingerprint image may be 1000 to 2000 dpi.
  • Resolution can be understood as the image resolution, which is the number of pixels contained in a unit of inch.
  • Dpi Dots Per Inch
  • the fingerprint sweat hole in the fingerprint image to be identified is interpolated in this embodiment. In order to improve the resolution of the fingerprint image to be recognized, the fingerprint sweat hole in the first fingerprint image after the resolution is improved can be detected more accurately.
  • the pore size of the fingerprint sweat pores is between 50 and 250 ⁇ m, wherein the number of sweat pores having a size between 100 and 200 ⁇ m accounts for more than 60% of the total number of sweat pores.
  • it is mainly considered to detect sweat pores in the size range of 100 to 200 ⁇ m.
  • a sweat hole having a size between 100 and 200 ⁇ m is not more than 4 pixels in a fingerprint image with a resolution of 500 dpi. When the sweat hole of 4 pixel points is directly detected in the low resolution image, the detected sweat hole area is insufficient for fingerprint recognition, that is, the accuracy of fingerprint recognition is low at this time.
  • the pixel points of the transition gray value are generated around the original sweat hole, so that the number of pixels of the sweat hole is increased.
  • the number of pixels of the sweat hole in the first fingerprint image is 4 to 16.
  • the image shown in Fig. 2b can be obtained.
  • the resolution of the fingerprint image in Fig. 2b is improved, and the sweat hole in the figure is significantly increased.
  • one of interpolation algorithms such as a bilinear interpolation algorithm and a bicubic interpolation algorithm may be used to interpolate the fingerprint image to be recognized.
  • an interpolation algorithm may be used according to the resolution of the preset first fingerprint image and the resolution of the fingerprint image to be recognized, and the fingerprint image to be recognized is interpolated to obtain a first fingerprint image.
  • the resolution of the first fingerprint image is greater than the resolution of the fingerprint image to be identified.
  • the size of the first fingerprint image may be determined according to the resolution of the resolution of the first fingerprint image and the resolution of the fingerprint image to be recognized, and the size of the fingerprint image to be identified, according to the size and first of the fingerprint image to be identified.
  • Determining the relationship between the sizes of the fingerprint images determining the coordinates (X, Y) of the coordinates (x, y) of each pixel in the first fingerprint image in the fingerprint image to be recognized, and determining the to-be-interacted algorithm according to the interpolation algorithm and the corresponding coordinates Identifying a target pixel point around the corresponding coordinate in the fingerprint image, and determining a pixel value of the pixel point in the first fingerprint image according to the pixel value of the target pixel point.
  • the size of the fingerprint image to be identified is w0*h0 (width*height), the resolution is 500 dpi, and the first fingerprint image is required to reach 1000 dpi, and the size of the first fingerprint image is 2w0*2h0.
  • the corresponding coordinates (X, Y) of the pixel points (x, y) in the fingerprint image to be recognized in the first fingerprint image are (x/2, y/2).
  • a pixel value of a pixel point (x, y) in the first fingerprint image is determined according to pixel values of the four target pixel points. Specifically, for four target pixel points, linear interpolation can be performed in two directions respectively to obtain two transition pixel points, and then the obtained two transition pixel points are linearly interpolated to obtain pixel points (x, y). Pixel values.
  • the bilinear interpolation algorithm is a commonly used interpolation method in two-dimensional space. Experiments show that the bilinear interpolation algorithm has higher computational efficiency. Therefore, the bilinear interpolation algorithm can improve the computational efficiency when interpolating the fingerprint image.
  • sixteen target pixel points in the rectangular grid around the corresponding coordinates (X, Y) in the fingerprint image to be identified may be determined, according to the pixel values of the sixteen target pixel points and the preset bicubic
  • the interpolation formula determines the pixel value of the pixel point (x, y) in the first fingerprint image.
  • the bicubic interpolation algorithm is also a commonly used interpolation method in two-dimensional space. Experiments show that the accuracy of the pixel values calculated by the bicubic interpolation algorithm is better. Therefore, the bi-cubic interpolation algorithm can improve the calculation accuracy when interpolating the fingerprint image.
  • the fingerprint image collected by such a fingerprint collection device has a trapezoidal shape. distortion. Therefore, before the interpolation of the fingerprint image to be recognized, the trapezoidal distortion of the fingerprint image to be recognized may be corrected, and the corrected image is interpolated to obtain a first fingerprint image. Alternatively, after the interpolation of the fingerprint image to be recognized, the trapezoidal distortion of the image after the interpolation may be corrected to obtain a first fingerprint image.
  • Step S103 Detect a sweat hole area in the first fingerprint image according to the preset sweat hole pixel feature.
  • the preset sweat hole pixel feature may be a number feature of the sweat hole pixel point, for example, the number of sweat hole pixel points is 4 to 16; or may be a distribution position feature of the sweat hole pixel point, for example, a sweat hole.
  • the pixel points are distributed on the fingerprint ridge line; or, there may be a feature that there is a large difference between the pixel values of the sweat hole region and the fingerprint ridge line region.
  • the sweat hole pixel feature can also be a combination of the above several features.
  • the preset sweat hole pixel feature may be: in the binarized image, the connected domain of the number of pixels in a preset number range is a sweat hole region.
  • the first fingerprint image when detecting the sweat hole region in the first fingerprint image, the first fingerprint image may be binarized to obtain a binarized image, and each connected domain in the binarized image is detected for detecting Each connected domain determines a connected domain whose number of pixels is within a preset number range as a sweat hole region in the first fingerprint image. It has been found through experiments that the detection efficiency is higher when the sweat region is detected by the above-mentioned connected domain, and the method is more suitable for application in the embedded fingerprint collection device.
  • the method is a method for adaptively selecting a binarization threshold, and the application effect is remarkable.
  • the first fingerprint image Before the first fingerprint image is binarized, the first fingerprint image may be pre-processed to improve the contrast of the image, so that the second fingerprint image can be more accurate when binarized.
  • the pre-processing may include operations such as equalization, filtering, segmentation, and enhancement.
  • FIG. 2c is a binarized image obtained by binarizing the first fingerprint image in FIG. 2b.
  • each connected domain in the binarized image it can be detected by an eight-neighbor connectivity algorithm.
  • eight adjacent pixel points around the pixel point P are pixel points above, below, left, right, upper left, upper right, lower left, and lower right of the pixel P.
  • the gray value of the white portion is set to 1, and the gray value of the black texture portion is zero. Traversing all the pixels in FIG. 2d, if there are pixels in the eight adjacent pixels around the pixel P having a gray value of 1 and the gray value is not 0, it indicates that the gray value is not 0.
  • the pixel is connected to the pixel P, and the pixel and the pixel P whose gradation value is not 0 are marked as a connected domain.
  • the number of pixels in each connected domain is accumulated when the connected domain is marked, and the number of pixels of each connected domain is obtained.
  • a connected domain having a number of pixels within a preset number range may be determined as a sweat hole region in the first fingerprint image.
  • the preset quantity range may be a value determined according to experience, and may be, for example, a range of 4-16, or 3-16, or 4-17. Experiments have shown that setting the preset number range to the above range can make the detection result of the sweat hole more stable.
  • Determining the sweat hole area in the first fingerprint image can be understood as determining the coordinate position of the sweat hole area in the first fingerprint image.
  • the wavelet transform when detecting the sweat hole region in the first fingerprint image, the wavelet transform may also be used.
  • Wavelet transform can fully highlight some aspects of the image through transformation, and can perform localized analysis on time or spatial frequency.
  • the image is gradually multi-scale refined by telescopic translation operation, and finally reaches the high-frequency time subdivision, low-frequency frequency.
  • Subdivision which adapts to the requirements of time-frequency signal analysis, so that it can focus on any detail of the image.
  • This method has higher accuracy in detecting the sweat hole area, and has higher requirements on the calculation performance of the device, and is more suitable for application in a high-configuration device.
  • Step S104 Determine a sweat hole feature of the first fingerprint image according to the detected sweat hole area.
  • the sweat hole feature may include at least one of coordinates of a center point of the sweat hole region, a size of the sweat hole region, a direction field of the fingerprint ridge line where the sweat hole region is located, a number of sweat hole regions, and a density of the sweat hole region.
  • the sweat hole feature may be determined according to all sweat hole regions in the first fingerprint image, and the sweat hole feature may also be determined according to a partial sweat hole region in the first fingerprint image.
  • the detected sweat hole feature may be a sweat hole feature obtained according to characteristics of each sweat hole region in the first fingerprint image, or may be a direct combination of features of each sweat hole region in the first fingerprint image.
  • Step S105 Identify the fingerprint in the fingerprint image to be recognized according to the sweat hole feature.
  • the method includes: matching the sweat feature and the preset sweat feature, and identifying the fingerprint in the fingerprint image according to the matching result.
  • the preset sweat hole feature may be a sweat hole feature obtained when the process of step S101 to step S104 is performed on the sample fingerprint image in advance, for example, the coordinates including the center point of the sweat hole region, the size of the sweat hole region, and the fingerprint ridge where the sweat hole region is located. a feature of at least one of the directional fields of the line; or may be obtained by counting sweat features of a plurality of different sample fingerprint images, for example, including at least one of the number of sweat regions and the density of the sweat regions. feature.
  • the embodiment can perform interpolation by using the fingerprint image to be recognized to obtain a fingerprint image with higher resolution, so that the sweat hole region can be detected more accurately from the fingerprint image; the fingerprint image can be determined according to the detected sweat hole region.
  • the characteristics of the sweat hole according to the characteristics of the sweat hole to identify the fingerprint image for fingerprint recognition. Since the sweat hole feature on the fingerprint ridge line is more complicated and more difficult to be copied than the detail point, it can improve the accuracy of fingerprint recognition when applied to fingerprint recognition.
  • the solution of the embodiment of the present application does not need to change the optical path of the embedded optical fingerprint collection device, and does not need to improve the hardware. Only the software part inside the chip can be improved to improve the accuracy of fingerprint recognition, so the solution is easier to implement. .
  • the embodiment shown in FIG. 1 is modified to obtain the embodiment shown in FIG. 3.
  • the embodiment is applied to an electronic device having a fingerprint recognition function, such as a fingerprint collection device.
  • the method includes the following steps S301 to S305:
  • Step S301 Acquire a fingerprint image to be identified.
  • Step S302 Interpolating the fingerprint image to be recognized to obtain a first fingerprint image.
  • Step S303 Detect the sweat hole area in the first fingerprint image according to the preset sweat hole pixel feature.
  • the steps S301 to S303 in the embodiment are the same as the steps S101 to S103 in the embodiment shown in FIG. 1 .
  • Step S304 Determine a sweat hole feature of the first fingerprint image according to the detected sweat hole area.
  • the sweat hole feature may include at least one of a number of sweat holes and a distribution density of sweat holes.
  • the step may be: determining a total number N of all sweat regions in the first fingerprint image; or, according to the total number of all pixel points in the first fingerprint image, M1 and all sweat regions in the first fingerprint image.
  • the total number of all sweat regions in the first fingerprint image may be determined, and the first fingerprint is determined according to the total number of all pixels in the first fingerprint image and the total number of all sweat regions in the first fingerprint image.
  • Step S305 Determine whether the fingerprint in the fingerprint image to be identified is a living fingerprint according to the sweat hole feature and the preset first sweat hole feature.
  • the preset first sweat hole feature may include a total number range [N1, N2] of all sweat hole regions in the fingerprint image, and/or a distribution density range [ ⁇ 1, ⁇ 2] of the sweat hole region in the fingerprint image.
  • the first sweat feature can be obtained by pre-calculating the sweat feature of a large number of different sample fingerprint images.
  • the step of determining whether the sweat hole feature is within a range corresponding to the preset first sweat hole feature and if yes, determining that the fingerprint in the fingerprint image to be recognized is a living fingerprint; if not, determining the fingerprint to be recognized The fingerprint in the image is not a live fingerprint.
  • the preset first sweat hole feature is a total number range of all sweat hole regions in the fingerprint image [90, 200], and if the determined sweat hole feature in the first fingerprint image is 150, the fingerprint image to be recognized is determined.
  • the fingerprint in the image is a living fingerprint. If the number of sweat holes in the first fingerprint image is 50, it is determined that the fingerprint in the fingerprint image to be recognized is not a living fingerprint.
  • the preset first sweat hole feature is a distribution density range of the sweat hole region in the fingerprint image [0.1, 0.2], and if the distribution density of the sweat hole region in the first fingerprint image is determined to be 0.15, the fingerprint to be recognized is determined.
  • the fingerprint in the image is a living fingerprint. If the distribution density of the sweat hole region in the first fingerprint image is 0.05, it is determined that the fingerprint in the fingerprint image to be recognized is not a living fingerprint.
  • the living body detection is a technique for detecting whether a sample to be authenticated has a vital feature in a biometric process in order to prevent a malicious person from using a forged biometric of another person for identity authentication.
  • the sweat hole information on the fingerprint belongs to the third feature of the fingerprint, like the minutiae information of the second layer feature, it can be used for identification; and the number of sweat holes on the fingerprint is very large and difficult to be copied. Therefore, the fingerprint sweat hole can be used for living body detection.
  • the embodiment can perform interpolation by using the fingerprint image to be recognized to obtain a fingerprint image with higher resolution, so that the sweat hole region can be detected more accurately from the fingerprint image; the fingerprint image can be determined according to the detected sweat hole region.
  • the characteristics of the sweat hole according to the characteristics of the sweat hole to identify the fingerprint image for living detection. Since the sweat hole feature on the fingerprint ridge line is more complicated and more difficult to be copied than the detail point, it can improve the accuracy of detection when applied to the living body detection.
  • the fingerprint in the fingerprint image to be identified is determined as the living fingerprint
  • the fingerprint may be further authenticated according to the feature of the minutiae point.
  • the above embodiment will be more secure than a method of matching fingerprints by only the minutiae.
  • the embodiment shown in FIG. 1 is modified to obtain the embodiment shown in FIG. 4, which is applied to an electronic device having a fingerprint recognition function, such as a fingerprint collection device.
  • the method includes the following steps S401 to S405:
  • Step S401 Acquire a fingerprint image to be identified.
  • Step S402 Interpolating the fingerprint image to be recognized to obtain a first fingerprint image.
  • Step S403 Detect the sweat hole area in the first fingerprint image according to the preset sweat hole pixel feature.
  • the steps S401 to S403 in the embodiment are the same as the steps S101 to S103 in the embodiment shown in FIG. 1 .
  • Step S404 Determine a sweat hole feature of the first fingerprint image according to the detected sweat hole area.
  • the sweat hole feature of the first fingerprint image may be determined as the sweat hole feature of the first fingerprint image: the center point coordinate of the sweat hole region, the size of the sweat hole region, and the direction field of the fingerprint ridge line where the sweat hole region is located.
  • the size of the sweat hole area can be understood as the number of pixels contained in the sweat hole area.
  • the directional field is a vector group that describes the shape and direction of the ridgeline.
  • the above sweat feature can also be referred to as a sweat index.
  • the features of all the sweat hole regions in the first fingerprint image may be used as the sweat hole feature of the first fingerprint image, and the features of the partial sweat region in the first fingerprint image may also be used as The sweat hole area of the first fingerprint image.
  • Step S405 Determine an object corresponding to the sweat hole feature as an object to which the fingerprint in the fingerprint image to be identified belongs according to the sweat hole feature, the pre-stored correspondence relationship between the second sweat hole feature and the object.
  • the second sweat hole feature may include at least one of a center point coordinate of the sweat hole region, a size of the sweat hole region, and a direction field of the fingerprint ridge line where the sweat hole region is located.
  • the second sweat hole feature may also be a relative feature quantity determined between the sweat hole area according to at least one of a center point coordinate of the sweat hole area, a size of the sweat hole area, and a direction field of the fingerprint ridge line where the sweat hole area is located. For example, the relative distance between the center points of the plurality of sweat hole regions, the rotation angle, and the like.
  • Corresponding relationship between the second sweat hole feature and the object stored in advance may be obtained by: acquiring a sample fingerprint image and a corresponding object in advance, and interpolating the sample fingerprint image, and detecting a sweat hole region of the image after interpolation, according to The detected sweat hole area determines the sweat hole characteristics, and generates a corresponding relationship between the sweat hole characteristics and the object.
  • determining an object to which a fingerprint belongs may be understood as identity authentication of the fingerprint.
  • the sweat hole feature may be matched with the second sweat hole feature in the corresponding relationship, and the object corresponding to the successfully matched second sweat hole feature may be the object corresponding to the sweat hole feature.
  • the second sweat hole is characterized by a relative feature quantity
  • a plurality of sweat hole regions in the sweat hole feature may be determined according to the sweat hole feature.
  • the relative feature quantity between the sweat hole features such as the relative distance between the center points of the plurality of sweat hole regions, the rotation angle, etc., according to the determined relative feature quantity between the sweat hole features, and the second sweat hole feature match.
  • Such a matching process can improve the matching accuracy.
  • the embodiment can perform interpolation by using the fingerprint image to be recognized to obtain a fingerprint image with higher resolution, so that the sweat hole region can be detected more accurately from the fingerprint image; the fingerprint image can be determined according to the detected sweat hole region.
  • the characteristics of the sweat hole according to the characteristics of the sweat hole to identify the fingerprint image for personal identification. Since the sweat hole feature on the fingerprint ridge line is more complicated and more difficult to be copied than the detail point, it can improve the accuracy of recognition when applied to fingerprint identification.
  • step S405 it may also be determined whether the fingerprint in the fingerprint image to be identified is a living fingerprint, and if so, step S405 is performed, that is, according to the above The sweat hole feature, the pre-stored second sweat hole feature and the object corresponding relationship, determine the object corresponding to the sweat hole feature, as the object of the fingerprint in the fingerprint image to be identified.
  • the method in the embodiment shown in FIG. 3 is used to determine whether the fingerprint in the fingerprint image to be identified is a living fingerprint. For details, refer to the embodiment shown in FIG. 3, and details are not described herein again.
  • the identity authentication of the fingerprint image before the identity authentication of the fingerprint image is to be performed, it is determined whether the fingerprint in the fingerprint image to be identified is a living fingerprint, and if so, the object to which the fingerprint belongs is identified, thereby improving the efficiency and accuracy of fingerprint recognition. .
  • step S404 the step of determining the sweat hole feature of the first fingerprint image according to the detected sweat hole region includes the following steps 1 to 3:
  • Step 1 Determine each detail point on the fingerprint ridgeline in the first fingerprint image.
  • the minutiae points include at least one of a bifurcation point, an end point, a ring, an island, and a bridge.
  • Step 2 Determine the target sweat hole area by using the sweat point area within the range of the preset distance as the center with each detail point as the center.
  • the preset distance may be a distance value determined in advance according to experience.
  • the preset distance can be a value in units of pixels.
  • the preset distance may be 6 to 10 pixels.
  • the range centered on the minutiae and radiused by the preset distance may be a circular area.
  • the step may be determined as a target sweat hole region by using one or more sweat hole regions in a range of a predetermined distance as a center centered on each detail point.
  • all the sweat hole regions in the range of the predetermined distance as the center and each of the detail points are determined as the target sweat hole regions. It is also possible to determine, as the target sweat hole region, a sweat hole region having a predetermined number of pixel points in a sweat hole region in a range of a predetermined distance as a center centered on each detail point.
  • the preset number may be an amount determined in advance according to experience, and may be, for example, 6, 7, 8, or 9.
  • the size of the sweat hole area refers to the total number of pixel points in the sweat hole area.
  • the 50 regions can be All the sweat hole areas are determined as the target sweat hole areas, and the sweat hole areas having a size of 8 pixels in the 50 areas may be determined as the target sweat hole areas.
  • FIG. 5a is an enlarged view of a partial image in the first fingerprint image of FIG. 2b, and FIG. 5a shows three sweats in a circular range with a radius of 7 pixels as a center centering on a certain bifurcation point. hole.
  • the sweat hole area where the three sweat holes in the circular range are located may be used as the target sweat hole area, or the sweat hole area in which the three sweat holes in the circular range are located may be A sweat hole area having a size of about 8 pixels is used as a target sweat hole area.
  • the details will be All the sweat hole areas within the range of the point-centered radius with the preset distance as the target are determined as the target sweat hole area.
  • the number of sweat holes in the range of more than a predetermined number of thresholds is more stable, so that the accuracy of fingerprint recognition can be improved.
  • a fingerprint image contains a total of 30 minutiae points
  • all the regions in each region can be calculated for 30 regions centered on 30 minutiae points and with a preset distance as a radius.
  • the number of sweat holes is summed to get 30 quantity sums. If 10 of the 30 quantity sums are greater than the preset number threshold, then the 10 numbers and all sweat areas in the corresponding area are determined as the target sweat area.
  • Step 3 Extract the feature of the target sweat region as the sweat hole feature of the first fingerprint image.
  • At least one of the center point coordinates of the target sweat region, the size of the target sweat region, and the direction field of the fingerprint ridge line where the target sweat region is located may be used as the sweat feature of the first fingerprint image.
  • the feature of all the sweat hole regions in the range of the radius of the preset point and the radius of the preset distance can be extracted as the sweat hole feature of the first fingerprint image, and it is not necessary to extract all the sweat in the first fingerprint image.
  • the characteristics of the hole area enable targeted selection of the sweat hole area, reduce the amount of calculation, and improve the processing efficiency.
  • the first fingerprint image when detecting the sweat hole region in the first fingerprint image according to the preset sweat hole pixel feature, the first fingerprint image is performed.
  • the above steps 1 To obtain a binarized fingerprint image, to detect each connected domain in the binarized fingerprint image, and to determine the connected domain of the number of pixels in the preset number range as the sweat hole region in the first fingerprint image, the above steps 1.
  • the step of determining a detail point on the fingerprint ridge line in the first fingerprint image may include steps 1a to 1c:
  • Step 1a Filling the binarized fingerprint image according to the sweat hole region in the first fingerprint image to obtain a filled image.
  • the step may specifically be: changing the sweat hole area in the first fingerprint image as the sweat hole area in the binarized fingerprint image, and replacing the pixel value of the pixel point of the sweat hole area in the binarized fingerprint image with the binary value.
  • Another pixel value in the fingerprint image For example, the binarized fingerprint image includes two pixel values of pixel values of 0 and 255. If the pixel value of the sweat region is 255, the pixel value of the pixel of the sweat region can be replaced with 0, that is, the padding is obtained. image.
  • the binarized fingerprint image includes two pixel values of 0 and 255, with 0 being black and 255 being white.
  • the black stripe pattern is the fingerprint ridge line part
  • the white point on the fingerprint ridge line is the sweat hole. Since the binarized fingerprint image 2c corresponds to the sweat hole region in the first fingerprint image (FIG. 2b), the sweat hole in the binarized fingerprint image can be determined according to the position of the sweat hole region in the first fingerprint image. In the area, the pixel value of the pixel point of the sweat hole area in FIG. 2c is replaced with 0, that is, the filling of the binarized fingerprint image is realized.
  • Step 1b Refine the filled image to obtain a refined image.
  • the padding image is refined, and may be performed by hitting or hitting the transform in the digital morphology, or may be performed by other algorithms, which is not specifically limited in this application.
  • a filled image can be obtained: there is no sweat hole area on the fingerprint ridge line in the image, that is, there is no small white point, and each fingerprint ridge line is a black line.
  • the filled image is refined, the skeleton of the fingerprint ridge line in the filled image can be extracted to obtain a line composed of single pixel points, and the image of the line composed of the single pixel points is a refined image.
  • FIG. 5b a schematic diagram of a refined image after refining the filled image.
  • This Fig. 5b is an image obtained by refining Fig. 2c and inverting the pixel values of the pixel points between 0 and 255.
  • Step 1c Determine a detail point on the fingerprint ridge line from the refined image as a detail point on the fingerprint ridge line in the first fingerprint image.
  • the minutiae of the fingerprint ridge line can be determined from the refined image according to the pixel characteristics of the minutiae point.
  • the bifurcation is characterized by having at least 3 pixel values around the pixel being the same as the pixel value of the pixel.
  • the feature of the endpoint is that there is one pixel value around the pixel and the pixel value of the pixel is the same.
  • each pixel point in the refinement image may be traversed, and when there is a pixel point that meets the above feature, the pixel point is determined as the minutiae point.
  • Figures 5c and 5d respectively indicate, by circles, the bifurcation points and endpoints determined from the refined image.
  • the binarized fingerprint image can be filled according to the sweat hole region, so that the refinement of the fill image can be made more accurate, thereby obtaining the detail point more accurately.
  • FIG. 6 is a schematic structural diagram of a fingerprint collection device according to an embodiment of the present disclosure.
  • the fingerprint collection device includes a sensor 601, a memory 602, and a processor 603. This device embodiment corresponds to the method embodiment shown in FIG.
  • the senor 601 is configured to collect a fingerprint image to be identified.
  • the processor 603 is configured to acquire a fingerprint image to be recognized collected by the sensor 601, perform interpolation on the fingerprint image to be recognized, obtain a first fingerprint image, and detect a sweat hole region in the first fingerprint image according to the preset sweat hole pixel feature, according to The detected sweat hole area determines the sweat hole feature of the first fingerprint image, and identifies the fingerprint in the fingerprint image to be recognized according to the sweat hole feature.
  • the resolution of the first fingerprint image is greater than the resolution of the fingerprint image to be identified.
  • the memory 602 can be used to store the fingerprint image to be recognized collected by the sensor 601.
  • the processor 603 can be configured to acquire a fingerprint image to be identified from the memory 602.
  • the memory 602 may include a random access memory (RAM), and may also include a non-volatile memory (NVM), such as at least one disk storage.
  • RAM random access memory
  • NVM non-volatile memory
  • the memory may also be at least one storage device located away from the aforementioned processor.
  • the processor 603 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), and the like; or a digital signal processing (DSP), dedicated integration.
  • CPU central processing unit
  • NP network processor
  • DSP digital signal processing
  • ASIC Application Specific Integrated Circuit
  • FPGA Field-Programmable Gate Array
  • the processor 603 is specifically configured to:
  • the fingerprint image to be recognized is interpolated by using one of a bilinear interpolation algorithm and a bicubic interpolation algorithm.
  • the processor 603 is specifically configured to:
  • FIG. 7 is another schematic structural diagram of a fingerprint collection device according to an embodiment of the present disclosure.
  • the fingerprint collection device includes a sensor 701, a memory 702, and a processor 703.
  • This embodiment is a modified embodiment based on the embodiment shown in FIG. 6.
  • the unmodified portion is the same as the embodiment shown in FIG. 6, and details are not described herein again.
  • the device embodiment corresponds to the method embodiment shown in FIG.
  • the sensor 701 and the memory 702 in this embodiment are the same as the sensor 601 and the memory 602 in the embodiment shown in FIG. 6, and details are not described herein again.
  • the processor 703 is configured to acquire a fingerprint image to be recognized collected by the sensor 701, perform interpolation on the fingerprint image to be recognized, obtain a first fingerprint image, and detect the first fingerprint image according to the preset sweat pixel feature.
  • the sweat hole region determines the sweat hole feature of the first fingerprint image according to the detected sweat hole region, and determines whether the fingerprint in the fingerprint image to be recognized is a living fingerprint according to the sweat hole feature and the preset first sweat hole feature.
  • the resolution of the first fingerprint image is greater than the resolution of the fingerprint image to be identified.
  • the sweat hole feature includes at least one of the number of sweat holes and the distribution density of the sweat holes.
  • FIG. 8 is another schematic structural diagram of a fingerprint collection device according to an embodiment of the present disclosure.
  • the fingerprint collection device includes a sensor 801, a memory 802, and a processor 803.
  • This embodiment is a modified embodiment based on the embodiment shown in FIG. 6.
  • the unmodified portion is the same as the embodiment shown in FIG. 6, and details are not described herein again.
  • the device embodiment corresponds to the method embodiment shown in FIG.
  • the sensor 801 and the memory 802 in this embodiment are the same as the sensor 601 and the memory 602 in the embodiment shown in FIG. 6, and details are not described herein again.
  • the processor 803 is configured to acquire a fingerprint image to be recognized collected by the sensor 801, perform interpolation on the fingerprint image to be recognized, obtain a first fingerprint image, and detect the first fingerprint image according to the preset sweat pixel feature. a sweat hole region, determining a sweat hole feature of the first fingerprint image according to the detected sweat hole region, and determining a corresponding sweat hole feature according to the sweat hole feature, the pre-stored second sweat hole feature and the object corresponding relationship
  • the object is the object to which the fingerprint in the fingerprint image to be identified belongs.
  • the resolution of the first fingerprint image is greater than the resolution of the fingerprint image to be identified.
  • the processor 803 is specifically configured to:
  • Determining the sweat feature of the first fingerprint image according to at least one of the following:
  • the processor 803 is specifically configured to:
  • Determining a detail point on the fingerprint ridge line in the first fingerprint image determining a target sweat hole area according to a sweat hole area centered on the detail point and having a radius of the preset distance, and extracting a feature of the target sweat hole area, as The sweat hole feature of the first fingerprint image.
  • the processor 803 is specifically configured to:
  • the hole area is determined as the target sweat hole area.
  • the processor 803 is specifically configured to:
  • the detecting the sweat hole region in the first fingerprint image according to the preset sweat hole pixel feature comprises: binarizing the first fingerprint image to obtain a binarized fingerprint image, and detecting each connected domain in the binarized fingerprint image
  • the binarized fingerprint image is filled according to the sweat hole region in the first fingerprint image to obtain a filled image. Fine-graining the filled image to obtain a refined image; determining a detail point on the fingerprint ridge line from the refined image as a detail point on the fingerprint ridgeline in the first fingerprint image.
  • the device embodiments shown in FIG. 6 to FIG. 8 are respectively obtained based on the method embodiments of FIG. 1 , FIG. 3 and FIG. 4 , and have the same technical effects as the corresponding methods. To avoid repetition, the technical effects of the device embodiments are I will not repeat them here.
  • the description is relatively simple, and the relevant parts can be referred to the description of the method embodiment.
  • the embodiment of the present application further provides a computer readable storage medium.
  • the computer readable storage medium stores a computer program.
  • the fingerprint identification method provided by the embodiment of the present application is implemented.
  • the method includes:
  • the resolution of the first fingerprint image is greater than the resolution of the fingerprint image to be identified
  • the fingerprint in the fingerprint image is to be identified.
  • the fingerprint image to be recognized may be interpolated to obtain a fingerprint image with a higher resolution, so that the sweat hole region can be detected more accurately from the fingerprint image; and the sweat hole feature of the fingerprint image can be determined according to the detected sweat hole region.
  • Fingerprint recognition is performed according to the characteristics of the sweat hole to identify the fingerprint image. Since the sweat hole feature on the fingerprint ridge line is more complicated and more difficult to be copied than the detail point, it can improve the accuracy of fingerprint recognition when applied to fingerprint recognition.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Collating Specific Patterns (AREA)
  • Image Input (AREA)

Abstract

本申请实施例提供了一种指纹识别方法及指纹采集设备。该方法包括：获取待识别指纹图像；对待识别指纹图像进行插值，得到第一指纹图像；其中，第一指纹图像的分辨率大于待识别指纹图像的分辨率；根据预设的汗孔像素特征，检测第一指纹图像中的汗孔区域；根据检测的汗孔区域，确定第一指纹图像的汗孔特征；根据上述汗孔特征，对待识别指纹图像中的指纹进行识别。应用本申请实施例提供的方案，能够提高指纹识别的准确性。

Description

一种指纹识别方法及指纹采集设备
本申请要求于2017年11月29日提交中国专利局、申请号为201711228455.9、发明名称为“一种指纹识别方法及指纹采集设备”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请设计图像处理技术领域，特别是涉及一种指纹识别方法及指纹采集设备。
背景技术
指纹识别可以应用在很多领域，例如门禁、设备解锁、支付等方面。指纹采集设备采集用户的指纹图像后，可以提取指纹图像中的细节点特征，例如指纹图像中指纹脊线上的分叉点和端点等特征，根据提取的细节点特征对指纹进行识别。
但是，用户指纹中的细节点信息很容易泄露。当恶意者收集到用户的指纹时，会利用指纹中的细节点特征伪造用户指纹。这种伪造的指纹成本很低，但却可以制作出相当清晰的指纹细节点特征。因此，上述指纹识别方法在识别真伪指纹时准确性不高，存在安全隐患。
发明内容
本申请实施例的目的在于提供了一种指纹识别方法及指纹采集设备，以提高指纹识别的准确性和安全性。具体的技术方案如下。
第一方面，本申请实施例提供了一种指纹识别方法。该方法包括：
获取待识别指纹图像；
对所述待识别指纹图像进行插值，得到第一指纹图像；其中，所述第一指纹图像的分辨率大于所述待识别指纹图像的分辨率；
根据预设的汗孔像素特征，检测所述第一指纹图像中的汗孔区域；
根据检测的汗孔区域，确定所述第一指纹图像的汗孔特征；
根据所述汗孔特征，对所述待识别指纹图像中的指纹进行识别。
可选的，所述对所述待识别指纹图像进行插值的步骤，包括：
采用双线性插值算法、双三次插值算法中的一种，对所述待识别指纹图像进行插值。
可选的，所述根据预设的汗孔像素特征，检测所述第一指纹图像中的汗孔区域的步骤，包括：
对所述第一指纹图像进行二值化，得到二值化图像；
检测所述二值化图像中的各个连通域；
针对所检测的各个连通域，将像素点数量在预设数量范围内的连通域确定为所述第一指纹图像中的汗孔区域。
可选的，所述根据所述汗孔特征，对所述待识别指纹图像中的指纹进行识别的步骤，包括：
根据所述汗孔特征和预设的第一汗孔特征，确定所述待识别指纹图像中的指纹是否为活体指纹。
可选的，所述汗孔特征包括：汗孔数量、汗孔的分布密度中的至少一种。
可选的，所述根据所述汗孔特征，对所述待识别指纹图像中的指纹进行识别的步骤，包括：
根据所述汗孔特征、预先存储的第二汗孔特征与对象的对应关系，确定所述汗孔特征对应的对象，作为所述待识别指纹图像中的指纹归属的对象。
可选的，所述根据检测的汗孔区域，确定所述第一指纹图像的汗孔特征的步骤，包括：
根据以下内容中的至少一种，确定所述第一指纹图像的汗孔特征：
汗孔区域的中心点坐标、汗孔区域的大小、汗孔区域所在指纹脊线的方向场。
可选的，所述根据检测的汗孔区域，确定所述第一指纹图像的汗孔特征的步骤，包括：
确定所述第一指纹图像中指纹脊线上的细节点；
以各个细节点为中心、以预设距离为半径的范围内的汗孔区域，确定目标汗孔区域；
提取所述目标汗孔区域的特征，作为所述第一指纹图像的汗孔特征。
可选的，所述以各个细节点为中心、以预设距离为半径的范围内的汗孔区域，确定目标汗孔区域的步骤，包括：
以各个细节点为中心、以预设距离为半径的范围内的一个或多个汗孔区域，确定为目标汗孔区域；或者，
针对每个细节点，若以所述细节点为中心、以预设距离为半径的范围内 的所有汗孔区域的数量和大于预设数量阈值，则将以所述细节点为中心、以预设距离为半径的范围内的所有汗孔区域确定为目标汗孔区域。
可选的，当根据预设的汗孔像素特征，检测所述第一指纹图像中的汗孔区域的步骤包括：对所述第一指纹图像进行二值化，得到二值化指纹图像，检测所述二值化指纹图像中的各个连通域，将像素点数量在预设数量范围内的连通域确定为所述第一指纹图像中的汗孔区域时，所述确定所述第一指纹图像中指纹脊线上的细节点的步骤，包括：
根据所述第一指纹图像中的汗孔区域，对所述二值化指纹图像进行填充，得到填充图像；
对所述填充图像进行细化，得到细化图像；
从所述细化图像中确定指纹脊线上的细节点，作为所述第一指纹图像中指纹脊线上的细节点。
第二方面，本申请实施例提供了一种指纹采集设备。该设备包括：传感器、处理器和存储器；
所述传感器，用于采集待识别指纹图像；
所述处理器，用于获取所述传感器采集的待识别指纹图像，对所述待识别指纹图像进行插值，得到第一指纹图像，根据预设的汗孔像素特征，检测所述第一指纹图像中的汗孔区域，根据检测的汗孔区域，确定所述第一指纹图像的汗孔特征，根据所述汗孔特征，对所述待识别指纹图像中的指纹进行识别；其中，所述第一指纹图像的分辨率大于所述待识别指纹图像的分辨率。
可选的，所述处理器具体用于：采用双线性插值算法、双三次插值算法中的一种，对所述待识别指纹图像进行插值。
可选的，所述处理器具体用于：
对所述第一指纹图像进行二值化，得到二值化图像；检测所述二值化图像中的各个连通域；针对所检测的各个连通域，将像素点数量在预设数量范围内的连通域确定为所述第一指纹图像中的汗孔区域。
可选的，所述处理器具体用于：根据所述汗孔特征和预设的第一汗孔特征，确定所述待识别指纹图像中的指纹是否为活体指纹。
可选的，所述汗孔特征包括：汗孔数量、汗孔的分布密度中的至少一种。
可选的，所述处理器具体用于：根据所述汗孔特征、预先存储的第二汗孔特征与对象的对应关系，确定所述汗孔特征对应的对象，作为所述待识别 指纹图像中的指纹归属的对象。
可选的，所述处理器具体用于根据以下内容中的至少一种，确定所述第一指纹图像的汗孔特征：汗孔区域的中心点坐标、汗孔区域的大小、汗孔区域所在指纹脊线的方向场。
可选的，所述处理器具体用于：
确定所述第一指纹图像中指纹脊线上的各个细节点，以各个细节点为中心、以预设距离为半径的范围内的汗孔区域，确定目标汗孔区域，提取所述目标汗孔区域的特征，作为所述第一指纹图像的汗孔特征。
可选的，所述处理器具体用于：
以各个细节点为中心、以预设距离为半径的范围内的一个或多个汗孔区域，确定为目标汗孔区域；或者，
针对每个细节点，若以所述细节点为中心、以预设距离为半径的范围内的所有汗孔区域的数量和大于预设数量阈值，则将以所述细节点为中心、以预设距离为半径的范围内的所有汗孔区域确定为目标汗孔区域。
可选的，所述处理器具体用于：
当根据预设的汗孔像素特征，检测所述第一指纹图像中的汗孔区域包括：对所述第一指纹图像进行二值化，得到二值化指纹图像，检测所述二值化指纹图像中的各个连通域，将像素点数量在预设数量范围内的连通域确定为所述第一指纹图像中的汗孔区域时，根据所述第一指纹图像中的汗孔区域，对所述二值化指纹图像进行填充，得到填充图像；对所述填充图像进行细化，得到细化图像；从所述细化图像中确定指纹脊线上的细节点，作为所述第一指纹图像中指纹脊线上的细节点。
第三方面，本申请实施例提供了一种计算机可读存储介质。该计算机可读存储介质内存储有计算机程序，所述计算机程序被处理器执行时实现本申请实施例提供的指纹识别方法。该方法包括：
获取待识别指纹图像；
对所述待识别指纹图像进行插值，得到第一指纹图像；其中，所述第一指纹图像的分辨率大于所述待识别指纹图像的分辨率；
根据预设的汗孔像素特征，检测所述第一指纹图像中的汗孔区域；
根据检测的汗孔区域，确定所述第一指纹图像的汗孔特征；
根据所述汗孔特征，对所述待识别指纹图像中的指纹进行识别。
本申请实施例提供的指纹识别方法及指纹采集设备，可以对待识别指纹图像进行插值得到分辨率比待识别指纹图像的分辨率大的第一指纹图像，根据预设的汗孔像素特征，检测第一指纹图像中的汗孔区域，并根据汗孔区域确定第一指纹图像的汗孔特征，根据该汗孔特征对待识别指纹图像中的指纹进行识别。本申请实施例，可以采用对待识别指纹图像进行插值，得到较高分辨率的指纹图像，这样能更准确地从指纹图像中检测汗孔区域；根据检测的汗孔区域可以确定指纹图像的汗孔特征，根据汗孔特征对待识别指纹图像进行指纹识别。由于指纹脊线上的汗孔特征相比于细节点而言，其特征更复杂，更难以被复制，因此应用于指纹识别时能够提高指纹识别的准确性。当然，实施本申请的任一产品或方法并不一定需要同时达到以上所述的所有优点。
附图说明
为了更清楚地说明本申请实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单的介绍。显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本申请实施例提供的指纹识别方法的一种流程示意图；
图2a为本申请实施例提供的待识别指纹图像示意图；
图2b为本申请实施例提供的插值后指纹图像示意图；
图2c为本申请实施例提供的二值化图像示意图；
图2d为本申请实施例提供的像素点周围的八个相邻像素点示意图；
图3为本申请实施例提供的指纹识别方法的另一种流程示意图；
图4为本申请实施例提供的指纹识别方法的再一种流程示意图；
图5a为本申请实施例提供的筛选汗孔区域的一种示意图；
图5b为本申请实施例提供的细化图像示意图；
图5c和图5d分别为本申请实施例提供的细化图像中分叉点和端点示意图；
图6～图8分别为本申请实施例提供的指纹采集设备的几种结构示意图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行 清楚、完整的描述。显然，所描述的实施例仅仅是本申请的一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动的前提下所获得的所有其他实施例，都属于本申请保护的范围。
在本申请实施例中，指纹采集设备可以对指纹图像进行指纹识别。指纹采集设备用以进行指纹识别的指纹图像为低分辨率的指纹图像，例如，分辨率介于500～1000dpi的指纹图像。当指纹采集设备对指纹图像进行指纹识别时，可以根据指纹图像中的细节点特征进行指纹识别。但是细节点特征容易被恶意者伪造，进而制作出假冒指纹，导致指纹采集设备对指纹识别的准确性和安全性不高。
为了提高指纹识别的准确性和安全性，本申请实施例提供了一种指纹识别方法及指纹采集设备。下面通过具体实施例，对本申请进行详细说明。
图1为本申请实施例提供的指纹识别方法的一种流程示意图。该方法应用于具有指纹识别功能的电子设备，例如指纹采集设备等。所述方法包括：
步骤S101：获取待识别指纹图像。
本实施例中，电子设备可以具有指纹采集功能，也可以不具有指纹采集功能。本步骤中，当电子设备具有指纹采集功能时，获取待识别指纹图像时，可以直接获取采集的待识别指纹图像。当电子设备不具有指纹采集功能时，可以从其他设备中获取待识别指纹图像。
其中，待识别指纹图像为低分辨率的指纹图像。本实施例中的待识别指纹图像，可以理解为包含指纹汗孔，但是无法直接从待识别指纹图像中检测到足以进行指纹识别的汗孔区域的指纹图像。指纹汗孔为皮肤表面的汗腺开口，一般位于皮肤表面的波状隆起部的顶端，属于指纹的第三层特征。皮肤表面的波状隆起部在指纹图像中即为指纹脊线。待识别指纹图像可以为分辨率处于预设分辨率范围内的指纹图像。分辨率处于预设分辨率范围内的指纹图像为低分辨率图像。预设分辨率范围可以为500～1000dpi或其他范围。例如，待识别指纹图像可以为分辨率介于500～1000dpi的指纹图像。低于500dpi的指纹图像一般认为几乎不存在汗孔信息，高于1000dpi的指纹图像需要由具有较高配置的指纹采集设备采集。其中，上述举例中用于描述低分辨率的指纹图像的分辨率下限500dpi和分辨率上限1000dpi只是一种举例，该分辨率下限也可以为499dpi、498dpi等值，该分辨率上限也可以为1001dpi、1002dpi 等值。本申请并不对上述分辨率下限和分辨率上限进行具体限定。
在一种具体的实施方式中，本实施例可以应用在常见的门禁指纹设备中。这种门禁指纹设备采集的指纹图像的分辨率通常在500dpi～1000dpi之间。
例如，图2a为500dpi的光学指纹采集设备采集的指纹图像，图中指纹脊线上的小白点即为汗孔。这种指纹图像中携带指纹汗孔，但是指纹图像中的汗孔的像素点不明显，无法直接用于检测或对比，此时可以执行步骤S102。
步骤S102：对待识别指纹图像进行插值，得到第一指纹图像。
其中，第一指纹图像的分辨率大于待识别指纹图像的分辨率。例如，待识别指纹图像的分辨率为500～1000dpi时，第一指纹图像的分辨率可以为1000～2000dpi。分辨率可以理解为图像分辨率，即表示单位英寸中所包含的像素点数量。dpi(Dots Per Inch，每英寸点数)是一个图像量度单位，指每一英寸长度中取样、可显示或输出点的数目。
由于待识别指纹图像的分辨率较低，无法从待识别指纹图像中直接检测指纹汗孔，因此为了能利用待识别指纹图像中的指纹汗孔的信息，本实施例对待识别指纹图像进行插值，以提高待识别指纹图像的分辨率，使提高分辨率后的第一指纹图像中的指纹汗孔能够被更准确地检测到。
通过对大量指纹样本的检测可知，指纹汗孔的孔径大小在50～250μm之间，其中，大小在100～200μm之间的汗孔数量占汗孔总数量的60％以上。本实施例可以主要考虑检测100～200μm这一尺寸范围内的汗孔。大小在100～200μm之间的汗孔在分辨率500dpi的指纹图像中大概不超过4个像素点。直接在低分辨率图像中检测4个像素点大小的汗孔时，所检测到的汗孔区域不足以进行指纹识别，即此时进行指纹识别时准确性较低。
在对待识别指纹图像进行插值之后，原始汗孔周边会产生过渡灰度值的像素点，从而使汗孔的像素点数量增加。例如，第一指纹图像中汗孔的像素点数量在4～16个。通过选择合适的处理过程和合适的阈值，能更有效地、更准确地从第一指纹图像中检测出放大后的汗孔。
例如，对图2a中的指纹图像进行插值后，可以得到图2b所示图像。通过对这两个图进行对比，可以看出，图2b中的指纹图像的分辨率提高了，图中的汗孔明显增大了。
对待识别指纹图像进行插值时，可以采用双线性插值算法、双三次插值算法等插值算法中的一种，对待识别指纹图像进行插值。
在对待识别指纹图像进行插值时，可以根据预先设置的第一指纹图像的分辨率和待识别指纹图像的分辨率，采用插值算法，对待识别指纹图像进行插值，得到第一指纹图像。其中，第一指纹图像的分辨率大于待识别指纹图像的分辨率。具体的，可以根据第一指纹图像的分辨率和待识别指纹图像的分辨率的大小关系，以及待识别指纹图像的尺寸，确定第一指纹图像的尺寸，根据待识别指纹图像的尺寸和第一指纹图像的尺寸之间的关系，确定第一指纹图像中每个像素点的坐标(x，y)在待识别指纹图像中的对应坐标(X，Y)，根据插值算法以及对应坐标，确定待识别指纹图像中该对应坐标周围的目标像素点，根据目标像素点的像素值，确定第一指纹图像中像素点的像素值。
例如，待识别指纹图像的尺寸为w0*h0(宽*高)，分辨率为500dpi，第一指纹图像要求达到1000dpi，则第一指纹图像的尺寸为2w0*2h0。第一指纹图像中像素点(x，y)在待识别指纹图像中的对应坐标(X，Y)为(x/2，y/2)。
采用双线性插值算法时，可以确定待识别指纹图像中对应坐标(X，Y)周围的四个目标像素点。根据这四个目标像素点的像素值，确定第一指纹图像中像素点(x，y)的像素值。具体的，针对四个目标像素点，可以分别在两个方向上进行线性插值，得到两个过渡像素点，再对得到的两个过渡像素点进行线性插值，得到像素点(x，y)的像素值。
双线性插值算法是二维空间中常用的插值方法。实验表明，双线性插值算法的计算效率较高。因此，采用双线性插值算法对待识别指纹图像进行插值时能够提高计算效率。
采用双三次插值算法时，可以确定待识别指纹图像中对应坐标(X，Y)周围矩形网格中的十六个目标像素点，根据这十六个目标像素点的像素值以及预设双三次插值公式，确定第一指纹图像中像素点(x，y)的像素值。
双三次插值算法也是二维空间中常用的插值方法。实验表明，采用双三次插值算法计算得到的像素值的精度较好。因此，采用双三次插值算法对待识别指纹图像进行插值时能够提高计算精度。
由于实际中很多指纹采集设备采用棱镜采集指纹图像，由于图像采集平面与感光元件的主光轴之间不是垂直的，而是存在一定的倾角，这使得这类指纹采集设备采集的指纹图像存在梯形畸变。因此，在对待识别指纹图像进行插值之前，还可以校正待识别指纹图像的梯形畸变，对校正后的图像进行插值，得到第一指纹图像。或者，也可以在对待识别指纹图像进行插值之后， 再校正插值之后图像的梯形畸变，得到第一指纹图像。
步骤S103：根据预设的汗孔像素特征，检测第一指纹图像中的汗孔区域。
其中，预设的汗孔像素特征可以为汗孔像素点的数量特征，例如，汗孔像素点的数量在4～16个等；也可以为汗孔像素点的分布位置特征，例如，汗孔像素点分布在指纹脊线上；或者，可以为汗孔区域与指纹脊线区域的像素值之间存在较大差异的特征。汗孔像素特征也可以为上述几种特征的结合。
在一种具体的实施方式中，预设的汗孔像素特征可以为：二值化图像中，像素点数量在预设数量范围内的连通域为汗孔区域。根据上述特征，本步骤在检测第一指纹图像中的汗孔区域时，可以对第一指纹图像进行二值化，得到二值化图像，检测二值化图像中的各个连通域，针对所检测的各个连通域，将像素点数量在预设数量范围内的连通域确定为第一指纹图像中的汗孔区域。经过实验发现，采用上述连通域的方式检测汗孔区域时，计算效率较高，该方式更适合应用在嵌入式的指纹采集设备中。
对第一指纹图像进行二值化时，可以采用最大类间方差法即大津法(Otsu)进行。该方法为一种自适应选取二值化阈值的方法，应用效果显著。
在对第一指纹图像进行二值化之前，可以对第一指纹图像进行预处理，提高图像的对比度，这样对第一指纹图像进行二值化时能够更准确。其中，预处理可以包括均衡化、滤波、分割和增强等操作。
例如，图2c为对图2b中的第一指纹图像进行二值化后得到的二值化图像。
检测二值化图像中的各个连通域时，可以采用八邻域连通算法检测。例如，参见图2d，像素点P周围的八个相邻的像素点为该像素点P的上方、下方、左侧、右侧、左上角、右上角、左下角、右下角的像素点。对于二值化图像来说，设置白色部分的灰度值为1，黑色纹路部分的灰度值为0。遍历图2d中的所有像素点，若一灰度值为1的像素点P周围的八个相邻像素点中存在灰度值不为0的像素点，则表明该灰度值不为0的像素点和像素点P连通，将该灰度值不为0的像素点和像素点P标记为连通域。在标记连通域时对每个连通域内的像素点的数量进行累加，得到每个连通域的像素点数量。
针对所检测的各个连通域，可以将像素点数量在预设数量范围内的连通域确定为第一指纹图像中的汗孔区域。其中，预设数量范围可以为根据经验确定的值，例如可以为4～16，或3～16，或4～17等数量范围。实验表明，将 预设数量范围设置为上述范围，可以使得汗孔的检测结果更加稳定。
确定第一指纹图像中的汗孔区域，可以理解为确定第一指纹图像中汗孔区域的坐标位置。
在一种具体的实施方式中，检测第一指纹图像中的汗孔区域时，也可以采用小波变换的方式进行。小波变换通过变换能够充分突出图像某些方面的特征，能对时间或空间频率进行局部化分析，通过伸缩平移运算对图像逐步进行多尺度细化，最终达到高频处时间细分，低频处频率细分，能自适应时频信号分析的要求，从而可聚焦到图像的任意细节。这种方法在检测汗孔区域时精度更高，同时对设备计算性能的要求也更高，更适合应用在高配置的设备中。
步骤S104：根据检测的汗孔区域，确定第一指纹图像的汗孔特征。
其中，汗孔特征可以包括汗孔区域中心点的坐标、汗孔区域的大小、汗孔区域所在的指纹脊线的方向场、汗孔区域的数量、汗孔区域的密度等中的至少一种。
在确定第一指纹图像的汗孔特征时，可以根据第一指纹图像中的所有汗孔区域来确定汗孔特征，也可以根据第一指纹图像中的部分汗孔区域来确定汗孔特征。
检测的汗孔特征可以为根据第一指纹图像中各个汗孔区域的特征得到的汗孔特征，也可以为第一指纹图像中各个汗孔区域的特征的直接组合。
步骤S105：根据上述汗孔特征，对待识别指纹图像中的指纹进行识别。
对待识别指纹图像中的指纹进行识别时，可以包括：将上述汗孔特征和预设汗孔特征进行匹配，根据匹配结果，对待识别指纹图像中的指纹进行识别。预设汗孔特征可以为预先对样本指纹图像执行步骤S101～步骤S104的过程时得到的汗孔特征，例如包括汗孔区域中心点的坐标、汗孔区域的大小、汗孔区域所在的指纹脊线的方向场中的至少一种的特征；也可以为对大量不同的样本指纹图像的汗孔特征进行统计后得到，例如包括汗孔区域的数量、汗孔区域的密度中的至少一种的特征。
由上述内容可知，本实施例可以采用对待识别指纹图像进行插值，得到较高分辨率的指纹图像，这样能更准确地从指纹图像中检测汗孔区域；根据检测的汗孔区域可以确定指纹图像的汗孔特征，根据汗孔特征对待识别指纹图像进行指纹识别。由于指纹脊线上的汗孔特征相比于细节点而言，其特征 更复杂，更难以被复制，因此应用于指纹识别时能够提高指纹识别的准确性。
应用本申请实施例的方案，无需改动嵌入式光学指纹采集设备的光路，也无需对硬件进行改进，只需改进芯片内部的软件部分，即可提高指纹识别的准确性，因此本方案较易于实施。
在本申请的另一实施例中，对图1所示实施例加以改进，可以得到图3所示实施例，该实施例应用于具有指纹识别功能的电子设备，例如指纹采集设备等。该方法包括以下步骤S301～步骤S305：
步骤S301：获取待识别指纹图像。
步骤S302：对待识别指纹图像进行插值，得到第一指纹图像。
步骤S303：根据预设的汗孔像素特征，检测第一指纹图像中的汗孔区域。
其中，本实施例中的步骤S301～步骤S303分别与图1所示实施例中的步骤S101～步骤S103相同，详细说明请参见图1所示实施例的相关部分，此处不再赘述。
步骤S304：根据检测的汗孔区域，确定第一指纹图像的汗孔特征。
其中，上述汗孔特征可以包括：汗孔数量、汗孔的分布密度中的至少一种。
本步骤具体可以为，确定第一指纹图像中所有汗孔区域的总数量N；或者，也可以为，根据第一指纹图像中的所有像素点总数量M1和第一指纹图像中所有汗孔区域的像素点的总数量M2，确定第一指纹图像中汗孔区域的分布密度ρ＝M2/M1。其中，所有汗孔区域的像素点的总数量M2＝S1*m1+S2*m2+…+Sn*mn，Sn为第n个汗孔区域，mn为第n个汗孔区域的像素数量。
本步骤也可以为，确定第一指纹图像中所有汗孔区域的总数量，根据第一指纹图像中的所有像素点总数量和第一指纹图像中所有汗孔区域的总数量，确定第一指纹图像中汗孔区域的分布密度。例如，第一指纹图像中汗孔区域的分布密度可以采用以下公式确定：ρ＝N/M1。
步骤S305：根据上述汗孔特征和预设的第一汗孔特征，确定待识别指纹图像中的指纹是否为活体指纹。
其中，预设的第一汗孔特征可以包括指纹图像中所有汗孔区域的总数量范围[N1，N2]，和/或，指纹图像中汗孔区域的分布密度范围[ρ1，ρ2]。第一汗孔特征可以为预先对大量不同的样本指纹图像的汗孔特征进行统计后得到。
本步骤具体可以为，判断上述汗孔特征是否处于预设的第一汗孔特征对应的范围内，如果是，则确定待识别指纹图像中的指纹为活体指纹；如果否，则确定待识别指纹图像中的指纹不为活体指纹。
例如，预设的第一汗孔特征为指纹图像中所有汗孔区域的总数量范围[90，200]，如果确定的第一指纹图像中的汗孔特征为150个，则确定待识别指纹图像中的指纹为活体指纹。如果第一指纹图像中的汗孔特征为50个，则确定待识别指纹图像中的指纹不为活体指纹。
又如，预设的第一汗孔特征为指纹图像中汗孔区域的分布密度范围[0.1，0.2]，如果确定的第一指纹图像中汗孔区域的分布密度为0.15，则确定待识别指纹图像中的指纹为活体指纹。如果第一指纹图像中汗孔区域的分布密度为0.05，则确定待识别指纹图像中的指纹不为活体指纹。
如果确定待识别指纹图像中的指纹为活体指纹，则说明进行指纹识别的对象不是伪造的指纹膜。该过程也可以称为活体检测。活体检测为：为了防止恶意者将伪造的他人生物特征用于身份认证，在生物特征识别过程中，针对待认证样本是否具有生命特征进行检测的技术。
由于指纹上的汗孔信息属于指纹的第三次特征，与第二层特征的细节点信息一样，均可以用于身份识别；并且，指纹上汗孔区域的数量非常多，并且难以被复制，因此可以利用指纹汗孔进行活体检测。
由上述内容可知，本实施例可以采用对待识别指纹图像进行插值，得到较高分辨率的指纹图像，这样能更准确地从指纹图像中检测汗孔区域；根据检测的汗孔区域可以确定指纹图像的汗孔特征，根据汗孔特征对待识别指纹图像进行活体检测。由于指纹脊线上的汗孔特征相比于细节点而言，其特征更复杂，更难以被复制，因此应用于活体检测时能够提高检测的准确性。
在上述实施例确定待识别指纹图像中的指纹为活体指纹之后，还可以进一步根据细节点特征匹配指纹，对指纹进行身份认证。上述实施例相比于仅仅依靠细节点匹配指纹的方法将更加安全。
在本申请的另一实施例中，对图1所示实施例加以改进，可以得到图4所示实施例，该实施例应用于具有指纹识别功能的电子设备，例如指纹采集设备等。该方法包括以下步骤S401～步骤S405：
步骤S401：获取待识别指纹图像。
步骤S402：对待识别指纹图像进行插值，得到第一指纹图像。
步骤S403：根据预设的汗孔像素特征，检测第一指纹图像中的汗孔区域。
其中，本实施例中的步骤S401～步骤S403分别与图1所示实施例中的步骤S101～步骤S103相同，详细说明请参见图1所示实施例的相关部分，此处不再赘述。
步骤S404：根据检测的汗孔区域，确定第一指纹图像的汗孔特征。
本步骤中，可以将以下内容中的至少一种，确定为第一指纹图像的汗孔特征：汗孔区域的中心点坐标、汗孔区域的大小、汗孔区域所在指纹脊线的方向场。汗孔区域的大小可以理解为汗孔区域包含的像素点数量。方向场是描述纹线脊线形状和走向的向量组。上述汗孔特征也可以称为汗孔索引。
在确定第一指纹图像的汗孔特征时，可以将第一指纹图像中所有汗孔区域的特征作为第一指纹图像的汗孔特征，也可以将第一指纹图像中部分汗孔区域的特征作为第一指纹图像的汗孔区域。
步骤S405：根据上述汗孔特征、预先存储的第二汗孔特征与对象的对应关系，确定上述汗孔特征对应的对象，作为待识别指纹图像中的指纹归属的对象。
其中，对象可以用于表示个人身份。第二汗孔特征可以包括：汗孔区域的中心点坐标、汗孔区域的大小、汗孔区域所在指纹脊线的方向场中的至少一种。第二汗孔特征也可以为根据汗孔区域的中心点坐标、汗孔区域的大小、汗孔区域所在指纹脊线的方向场中的至少一种，确定的汗孔区域之间的相对特征量，例如多个汗孔区域的中心点之间的相对距离、旋转角度等。
预先存储的第二汗孔特征与对象的对应关系，可以为采用以下方式获得的：预先采集样本指纹图像和对应的对象，并对样本指纹图像进行插值，检测插值后图像的汗孔区域，根据检测的汗孔区域确定汗孔特征，生成汗孔特征与对象的对应关系。
本实施例中，确定指纹归属的对象，可以理解为对指纹进行身份认证。
本步骤具体可以为，将上述汗孔特征与上述对应关系中的第二汗孔特征进行匹配，将匹配成功的第二汗孔特征对应的对象作为上述汗孔特征对应的对象。当第二汗孔特征为相对特征量时，将上述汗孔特征与上述对应关系中的第二汗孔特征进行匹配时，可以根据上述汗孔特征，确定该汗孔特征中多个汗孔区域的汗孔特征之间的相对特征量，例如多个汗孔区域的中心点之间的相对距离、旋转角度等，根据确定的汗孔特征之间的相对特征量，与第二 汗孔特征进行匹配。这样的匹配过程能够提高匹配准确度。
由上述内容可知，本实施例可以采用对待识别指纹图像进行插值，得到较高分辨率的指纹图像，这样能更准确地从指纹图像中检测汗孔区域；根据检测的汗孔区域可以确定指纹图像的汗孔特征，根据汗孔特征对待识别指纹图像进行个人身份识别。由于指纹脊线上的汗孔特征相比于细节点而言，其特征更复杂，更难以被复制，因此应用于指纹身份识别时能够提高识别的准确性。
在本申请的另一实施例中，在图4所示实施例中，在步骤S405之前，还可以确定待识别指纹图像中的指纹是否为活体指纹，如果是，则执行步骤S405，即根据上述汗孔特征、预先存储的第二汗孔特征与对象的对应关系，确定上述汗孔特征对应的对象，作为待识别指纹图像中的指纹归属的对象。
在确定待识别指纹图像中的指纹是否为活体指纹时，可以采用图3所示实施例中的方式进行，具体说明参见图3所示实施例，此处不再赘述。
可见，本实施例可以在对待识别指纹图像进行身份认证之前，先确定待识别指纹图像中的指纹是否为活体指纹，如果是，再识别指纹归属的对象，这样能够提高指纹识别的效率和准确性。
由于第一指纹图像中的汗孔区域非常多，通常可以达到上百个，若将全部汗孔区域用于身份认证时的匹配计算，则计算过程非常耗时且没有必要。因此，可以有针对性地选取适量汗孔。
在本申请的另一实施例中，在图4所示实施例中，步骤S404，根据检测的汗孔区域，确定第一指纹图像的汗孔特征的步骤，包括以下步骤1～步骤3：
步骤1：确定第一指纹图像中指纹脊线上的各个细节点。该细节点包括分叉点、端点、环、岛、桥中的至少一种。
步骤2：以各个细节点为中心、以预设距离为半径的范围内的汗孔区域，确定目标汗孔区域。
其中，预设距离可以为预先根据经验确定的距离数值。预设距离可以为以像素点个数为单位的数值。例如，预设距离可以为6～10个像素点。以细节点为中心、以预设距离为半径的范围，可以为圆形区域。
本步骤具体可以为，以各个细节点为中心、以预设距离为半径的范围内的一个或多个汗孔区域，确定为目标汗孔区域。
上述实施方式具体可以为，将以各个细节点为中心、以预设距离为半径 的范围内的所有汗孔区域确定为目标汗孔区域。也可以为，将以各个细节点为中心、以预设距离为半径的范围内的汗孔区域中，大小为预设数量个像素点的汗孔区域，确定为目标汗孔区域。预设数量可以为预先根据经验确定的数量，例如可以为6、7、8、9个。其中，汗孔区域的大小是指汗孔区域的像素点的总数量。
例如，某个指纹图像中共包含50个细节点，则在确定目标汗孔区域时，针对以50个细节点为中心，以预设距离为半径的50个区域，可以将该50个区域内的所有汗孔区域确定为目标汗孔区域，也可以将该50个区域内大小为8个像素点的汗孔区域确定为目标汗孔区域。
又如，图5a为图2b中第一指纹图像中局部图像的放大图，图5a中标示了以某一分叉点为中心，以7个像素点为半径的圆形范围内的三个汗孔。在确定目标汗孔区域时，可以将圆形范围内的三个汗孔所在的汗孔区域均作为目标汗孔区域，也可以将圆形范围内的三个汗孔所在的汗孔区域中，大小为8个像素点左右的汗孔区域作为目标汗孔区域。
在另一种具体实施方式中，针对每个细节点，若以各个细节点为中心、以预设距离为半径的范围内的所有汗孔区域的数量和大于预设数量阈值，则将以细节点为中心、以预设距离为半径的范围内的所有汗孔区域确定为目标汗孔区域。在这种实施方式中，数量和大于预设数量阈值的范围内的汗孔区域更稳定，从而可以提高指纹识别的准确性。
例如，某个指纹图像中共包含30个细节点，则在确定目标汗孔区域时，针对以30个细节点为中心、以预设距离为半径的30个区域，可以计算每个区域内的所有汗孔区域的数量和，得到30个数量和。如果这30个数量和中有10个数量和大于预设数量阈值，则将这10个数量和对应区域内的所有汗孔区域，确定为目标汗孔区域。
步骤3：提取目标汗孔区域的特征，作为第一指纹图像的汗孔特征。
例如，可以将目标汗孔区域的中心点坐标、目标汗孔区域的大小、目标汗孔区域所在指纹脊线的方向场中的至少一种作为第一指纹图像的汗孔特征。
可见，本实施例可以提取以细节点为中心、以预设距离为半径的范围内的所有汗孔区域的特征，作为第一指纹图像的汗孔特征，无需提取第一指纹图像中的所有汗孔区域的特征，能够有针对性地选择汗孔区域，减少计算量， 提高处理效率。
在本申请的另一实施例中，在图4所示实施例中，当根据预设的汗孔像素特征，检测第一指纹图像中的汗孔区域的步骤包括：对第一指纹图像进行二值化，得到二值化指纹图像，检测二值化指纹图像中的各个连通域，将像素点数量在预设数量范围内的连通域确定为第一指纹图像中的汗孔区域时，上述步骤1，确定第一指纹图像中指纹脊线上的细节点的步骤，可以包括步骤1a～步骤1c：
步骤1a：根据第一指纹图像中的汗孔区域，对二值化指纹图像进行填充，得到填充图像。
本步骤具体可以为，将第一指纹图像中的汗孔区域，作为二值化指纹图像中的汗孔区域，将二值化指纹图像中汗孔区域的像素点的像素值更换为该二值化指纹图像中的另一种像素值。例如，二值化指纹图像中包括像素值为0和255的两种像素值，如果汗孔区域的像素值为255，则可以将汗孔区域的像素点的像素值更换为0，即得到填充图像。
参见图2c所示二值化指纹图像，指纹脊线上存在很多小白点，这些小白点即为指纹脊线上的汗孔，这些区域会影响对图像的细化的准确性，因此可以对汗孔区域进行填充，以便对图像细化时准确性更好。
在图2c中，二值化指纹图像中包括0和255的两种像素值，0为黑色，255为白色。黑色的条状纹路为指纹脊线部分，指纹脊线上的小白点为汗孔。由于二值化指纹图像2c与第一指纹图像(图2b)中的汗孔区域是对应的，因此可以根据第一指纹图像中的汗孔区域的位置，确定二值化指纹图像中的汗孔区域，将图2c中的汗孔区域的像素点的像素值更换为0，即实现对二值化指纹图像的填充。
步骤1b：对填充图像进行细化，得到细化图像。
本步骤中，对填充图像进行细化，可以采用数字形态学中击中或击不中变换进行，也可以采用其他算法进行，本申请对此不做具体限定。
在对图2c进行填充之后，可以得到这样的填充图像：图像中指纹脊线上没有汗孔区域，即没有小白点，每条指纹脊线都是黑色的线条。对该填充图像进行细化时，可以提取填充图像中指纹脊线的骨架，得到单像素点组成的线条，这些单像素点组成的线条所在的图像即为细化图像。
参见图5b，为对填充图像进行细化后的细化图像示意图。该图5b为对图 2c进行细化且将像素点的像素值在0和255之间反转之后得到的图像。
步骤1c：从细化图像中确定指纹脊线上的细节点，作为第一指纹图像中指纹脊线上的细节点。
本步骤中，可以根据细节点像素特征，从细化图像中确定指纹脊线上的细节点。例如，针对细化图像中位于前景的像素点(非背景中的像素点)，分叉点的特征为：在该像素点周围存在至少3个像素值与该像素点的像素值相同。端点的特征为：在像素点周围存在1个像素值与该像素点的像素值相同。
从细化图像中确定指纹脊线上的细节点时，具体可以遍历细化图像中的每个像素点，当存在符合上述特征的像素点时，将该像素点确定为细节点。例如，图5c和图5d分别用圆圈标示出了从细化图像中确定的分叉点和端点。
综上，本实施例中可以根据汗孔区域对二值化指纹图像进行填充，从而可以使得对填充图像的细化更准确，从而更准确地得到细节点。
图6为本申请实施例提供的指纹采集设备的一种结构示意图。该指纹采集设备包括：传感器601、存储器602和处理器603。该设备实施例与图1所示方法实施例相对应。
在本实施例中，传感器601，用于采集待识别指纹图像。
处理器603，用于获取传感器601采集的待识别指纹图像，对待识别指纹图像进行插值，得到第一指纹图像，根据预设的汗孔像素特征，检测第一指纹图像中的汗孔区域，根据检测的汗孔区域，确定第一指纹图像的汗孔特征，根据汗孔特征，对待识别指纹图像中的指纹进行识别。其中，第一指纹图像的分辨率大于待识别指纹图像的分辨率。
在一种具体实施方式中，存储器602，可以用于存储传感器601采集的待识别指纹图像。处理器603可以用于从存储器602中获取待识别指纹图像。
其中，存储器602可以包括随机存取存储器(Random Access Memory，RAM)，也可以包括非易失性存储器(Non-Volatile Memory，NVM)，例如至少一个磁盘存储器。可选的，存储器还可以是至少一个位于远离前述处理器的存储装置。
上述处理器603可以是通用处理器，包括中央处理器(Central Processing Unit，CPU)、网络处理器(Network Processor，NP)等；还可以是数字信号处理器(Digital Signal Processing，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现场可编程门阵列(Field-Programmable Gate Array， FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。
在本申请的另一实施例中，在图6所示实施例中，处理器603具体用于：
采用双线性插值算法、双三次插值算法中的一种，对待识别指纹图像进行插值。
在本申请的另一实施例中，在图6所示实施例中，处理器603具体用于：
对第一指纹图像进行二值化，得到二值化图像；检测二值化图像中的各个连通域；针对所检测的各个连通域，将像素点数量在预设数量范围内的连通域确定为第一指纹图像中的汗孔区域。
图7为本申请实施例提供的指纹采集设备的另一结构示意图。该指纹采集设备包括：传感器701、存储器702和处理器703。该实施例为在图6所示实施例基础上的改进实施例，未改进之处与图6所示实施例相同，此处不再赘述。本设备实施例与图3所示方法实施例相对应。本实施例中的传感器701和存储器702分别与图6所示实施例中的传感器601和存储器602相同，具体内容不再赘述。
本实施例中，处理器703，用于获取传感器701采集的待识别指纹图像，对待识别指纹图像进行插值，得到第一指纹图像，根据预设的汗孔像素特征，检测第一指纹图像中的汗孔区域，根据检测的汗孔区域，确定第一指纹图像的汗孔特征，根据汗孔特征和预设的第一汗孔特征，确定待识别指纹图像中的指纹是否为活体指纹。其中，第一指纹图像的分辨率大于待识别指纹图像的分辨率。
在本申请的另一实施例中，在图7所示实施例中，汗孔特征包括：汗孔数量、汗孔的分布密度中的至少一种。
图8为本申请实施例提供的指纹采集设备的另一结构示意图。该指纹采集设备包括：传感器801、存储器802和处理器803。该实施例为在图6所示实施例基础上的改进实施例，未改进之处与图6所示实施例相同，此处不再赘述。本设备实施例与图4所示方法实施例相对应。本实施例中的传感器801和存储器802分别与图6所示实施例中的传感器601和存储器602相同，具体内容不再赘述。
本实施例中，处理器803，用于获取传感器801采集的待识别指纹图像，对待识别指纹图像进行插值，得到第一指纹图像，根据预设的汗孔像素特征， 检测第一指纹图像中的汗孔区域，根据检测的汗孔区域，确定第一指纹图像的汗孔特征，根据所述汗孔特征、预先存储的第二汗孔特征与对象的对应关系，确定所述汗孔特征对应的对象，作为待识别指纹图像中的指纹归属的对象。其中，第一指纹图像的分辨率大于待识别指纹图像的分辨率。
在本申请的另一实施例中，在图8所示实施例中，处理器803具体用于：
根据以下内容中的至少一种，确定第一指纹图像的汗孔特征：
汗孔区域的中心点坐标、汗孔区域的大小、汗孔区域所在指纹脊线的方向场。
在本申请的另一实施例中，在图8所示实施例中，处理器803具体用于：
确定第一指纹图像中指纹脊线上的细节点，根据以细节点为中心、以预设距离为半径的范围内的汗孔区域，确定目标汗孔区域，提取目标汗孔区域的特征，作为第一指纹图像的汗孔特征。
在本申请的另一实施例中，在图8所示实施例中，处理器803具体用于：
若以细节点为中心、以预设距离为半径的范围内的所有汗孔区域的数量和大于预设数量阈值，则将以细节点为中心、以预设距离为半径的范围内的所有汗孔区域确定为目标汗孔区域。
在本申请的另一实施例中，在图8所示实施例中，处理器803具体用于：
当根据预设的汗孔像素特征，检测第一指纹图像中的汗孔区域包括：对第一指纹图像进行二值化，得到二值化指纹图像，检测二值化指纹图像中的各个连通域，将像素点数量在预设数量范围内的连通域确定为第一指纹图像中的汗孔区域时，根据第一指纹图像中的汗孔区域，对二值化指纹图像进行填充，得到填充图像；对填充图像进行细化，得到细化图像；从细化图像中确定指纹脊线上的细节点，作为第一指纹图像中指纹脊线上的细节点。
由于上述图6～图8所示设备实施例分别是基于图1、图3和图4的方法实施例得到的，与对应的方法具有相同的技术效果，为避免重复，设备实施例的技术效果在此不再赘述。对于设备实施例而言，由于其基本相似于方法实施例，所以描述得比较简单，相关之处参见方法实施例的部分说明即可。
本申请实施例还提供了一种计算机可读存储介质，该计算机可读存储介质内存储有计算机程序，计算机程序被处理器执行时实现本申请实施例提供的指纹识别方法。该方法包括：
获取待识别指纹图像；
对待识别指纹图像进行插值，得到第一指纹图像；其中，第一指纹图像的分辨率大于待识别指纹图像的分辨率；
根据预设的汗孔像素特征，检测第一指纹图像中的汗孔区域；
根据检测的汗孔区域，确定第一指纹图像的汗孔特征；
根据所述汗孔特征，对待识别指纹图像中的指纹进行识别。
本实施例可以采用对待识别指纹图像进行插值，得到较高分辨率的指纹图像，这样能更准确地从指纹图像中检测汗孔区域；根据检测的汗孔区域可以确定指纹图像的汗孔特征，根据汗孔特征对待识别指纹图像进行指纹识别。由于指纹脊线上的汗孔特征相比于细节点而言，其特征更复杂，更难以被复制，因此应用于指纹识别时能够提高指纹识别的准确性。
需要说明的是，在本文中，诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且，术语“包括”、“包含”或者任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
本说明书中的各个实施例均采用相关的方式描述，各个实施例之间相同相似的部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。
以上所述仅为本申请的较佳实施例而已，并非用于限定本申请的保护范围。凡在本申请的精神和原则之内所做的任何修改、等同替换、改进等，均包含在本申请的保护范围内。

Claims (21)

  1. 一种指纹识别方法，其特征在于，所述方法包括：
    获取待识别指纹图像；
    对所述待识别指纹图像进行插值，得到第一指纹图像；其中，所述第一指纹图像的分辨率大于所述待识别指纹图像的分辨率；
    根据预设的汗孔像素特征，检测所述第一指纹图像中的汗孔区域；
    根据检测的汗孔区域，确定所述第一指纹图像的汗孔特征；
    根据所述汗孔特征，对所述待识别指纹图像中的指纹进行识别。
  2. 根据权利要求1所述的方法，其特征在于，所述对所述待识别指纹图像进行插值的步骤，包括：
    采用双线性插值算法、双三次插值算法中的一种，对所述待识别指纹图像进行插值。
  3. 根据权利要求1所述的方法，其特征在于，所述根据预设的汗孔像素特征，检测所述第一指纹图像中的汗孔区域的步骤，包括：
    对所述第一指纹图像进行二值化，得到二值化图像；
    检测所述二值化图像中的各个连通域；
    针对所检测的各个连通域，将像素点数量在预设数量范围内的连通域确定为所述第一指纹图像中的汗孔区域。
  4. 根据权利要求1～3任一项所述的方法，其特征在于，所述根据所述汗孔特征，对所述待识别指纹图像中的指纹进行识别的步骤，包括：
    根据所述汗孔特征和预设的第一汗孔特征，确定所述待识别指纹图像中的指纹是否为活体指纹。
  5. 根据权利要求4所述的方法，其特征在于，所述汗孔特征包括：汗孔数量、汗孔的分布密度中的至少一种。
  6. 根据权利要求1～3任一项所述的方法，其特征在于，所述根据所述汗孔特征，对所述待识别指纹图像中的指纹进行识别的步骤，包括：
    根据所述汗孔特征、预先存储的第二汗孔特征与对象的对应关系，确定所述汗孔特征对应的对象，作为所述待识别指纹图像中的指纹归属的对象。
  7. 根据权利要求6所述的方法，其特征在于，所述根据检测的汗孔区域，确定所述第一指纹图像的汗孔特征的步骤，包括：
    根据以下内容中的至少一种，确定所述第一指纹图像的汗孔特征：
    汗孔区域的中心点坐标、汗孔区域的大小、汗孔区域所在指纹脊线的方向场。
  8. 根据权利要求6所述的方法，其特征在于，所述根据检测的汗孔区域，确定所述第一指纹图像的汗孔特征的步骤，包括：
    确定所述第一指纹图像中指纹脊线上的各个细节点；
    以各个细节点为中心、以预设距离为半径的范围内的汗孔区域，确定目标汗孔区域；
    提取所述目标汗孔区域的特征，作为所述第一指纹图像的汗孔特征。
  9. 根据权利要求8所述的方法，其特征在于，所述以各个细节点为中心、以预设距离为半径的范围内的汗孔区域，确定目标汗孔区域的步骤，包括：
    以各个细节点为中心、以预设距离为半径的范围内的一个或多个汗孔区域，确定为目标汗孔区域；或者，
    针对每个细节点，若以所述细节点为中心、以预设距离为半径的范围内的所有汗孔区域的数量和大于预设数量阈值，则将以所述细节点为中心、以预设距离为半径的范围内的所有汗孔区域确定为目标汗孔区域。
  10. 根据权利要求8所述的方法，其特征在于，当根据预设的汗孔像素特征，检测所述第一指纹图像中的汗孔区域的步骤包括：对所述第一指纹图像进行二值化，得到二值化指纹图像，检测所述二值化指纹图像中的各个连通域，将像素点数量在预设数量范围内的连通域确定为所述第一指纹图像中的汗孔区域时，所述确定所述第一指纹图像中指纹脊线上的细节点的步骤，包括：
    根据所述第一指纹图像中的汗孔区域，对所述二值化指纹图像进行填充，得到填充图像；
    对所述填充图像进行细化，得到细化图像；
    从所述细化图像中确定指纹脊线上的细节点，作为所述第一指纹图像中指纹脊线上的细节点。
  11. 一种指纹采集设备，其特征在于，包括：传感器、处理器和存储器；
    所述传感器，用于采集待识别指纹图像；
    所述处理器，用于获取所述传感器采集的待识别指纹图像，对所述待识别指纹图像进行插值，得到第一指纹图像，根据预设的汗孔像素特征，检测所述第一指纹图像中的汗孔区域，根据检测的汗孔区域，确定所述第一指纹图像的汗孔特征，根据所述汗孔特征，对所述待识别指纹图像中的指纹进行识别；其中，所述第一指纹图像的分辨率大于所述待识别指纹图像的分辨率。
  12. 根据权利要求11所述的设备，其特征在于，所述处理器，具体用于：
    采用双线性插值算法、双三次插值算法中的一种，对所述待识别指纹图像进行插值。
  13. 根据权利要求11所述的设备，其特征在于，所述处理器，具体用于：
    对所述第一指纹图像进行二值化，得到二值化图像；检测所述二值化图像中的各个连通域；针对所检测的各个连通域，将像素点数量在预设数量范围内的连通域确定为所述第一指纹图像中的汗孔区域。
  14. 根据权利要求11～13任一项所述的设备，其特征在于，所述处理器，具体用于：
    根据所述汗孔特征和预设的第一汗孔特征，确定所述待识别指纹图像中的指纹是否为活体指纹。
  15. 根据权利要求14所述的设备，其特征在于，所述汗孔特征包括：汗孔数量、汗孔的分布密度中的至少一种。
  16. 根据权利要求11～13任一项所述的设备，其特征在于，所述处理器，具体用于：
    根据所述汗孔特征、预先存储的第二汗孔特征与对象的对应关系，确定所述汗孔特征对应的对象，作为所述待识别指纹图像中的指纹归属的对象。
  17. 根据权利要求16所述的设备，其特征在于，所述处理器，具体用于：
    根据以下内容中的至少一种，确定所述第一指纹图像的汗孔特征：
    汗孔区域的中心点坐标、汗孔区域的大小、汗孔区域所在指纹脊线的方向场。
  18. 根据权利要求16所述的设备，其特征在于，所述处理器，具体用于：
    确定所述第一指纹图像中指纹脊线上的各个细节点，以各个细节点为中心、以预设距离为半径的范围内的汗孔区域，确定目标汗孔区域，提取所述目标汗孔区域的特征，作为所述第一指纹图像的汗孔特征。
  19. 根据权利要求18所述的设备，其特征在于，所述处理器，具体用于：
    以各个细节点为中心、以预设距离为半径的范围内的一个或多个汗孔区域，确定为目标汗孔区域；或者，
    针对每个细节点，若以所述细节点为中心、以预设距离为半径的范围内的所有汗孔区域的数量和大于预设数量阈值，则将以所述细节点为中心、以预设距离为半径的范围内的所有汗孔区域确定为目标汗孔区域。
  20. 根据权利要求18所述的设备，其特征在于，所述处理器，具体用于：
    当根据预设的汗孔像素特征，检测所述第一指纹图像中的汗孔区域包括：对所述第一指纹图像进行二值化，得到二值化指纹图像，检测所述二值化指纹图像中的各个连通域，将像素点数量在预设数量范围内的连通域确定为所述第一指纹图像中的汗孔区域时，根据所述第一指纹图像中的汗孔区域，对所述二值化指纹图像进行填充，得到填充图像；对所述填充图像进行细化，得到细化图像；从所述细化图像中确定指纹脊线上的细节点，作为所述第一指纹图像中指纹脊线上的细节点。
  21. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质内存储有计算机程序，所述计算机程序被处理器执行时实现权利要求1-10任一所述的方法步骤。
PCT/CN2018/083528 2017-11-29 2018-04-18 一种指纹识别方法及指纹采集设备 WO2019104932A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711228455.9 2017-11-29
CN201711228455.9A CN109840458A (zh) 2017-11-29 2017-11-29 一种指纹识别方法及指纹采集设备

Publications (1)

Publication Number Publication Date
WO2019104932A1 true WO2019104932A1 (zh) 2019-06-06

Family

ID=66665367

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/083528 WO2019104932A1 (zh) 2017-11-29 2018-04-18 一种指纹识别方法及指纹采集设备

Country Status (2)

Country Link
CN (1) CN109840458A (zh)
WO (1) WO2019104932A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118171261A (zh) * 2023-10-09 2024-06-11 深圳市氿博传感科技有限公司 安防密码验证方法、装置、设备、介质及其计算机产品

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110633384B (zh) * 2019-09-19 2022-05-17 哈尔滨工业大学（深圳） 基于汗孔和多图匹配的高分辨率指纹检索方法、装置、系统及存储介质
CN111079626B (zh) * 2019-12-11 2023-08-01 深圳市迪安杰智能识别科技有限公司 一种活体指纹识别方法、电子设备及计算机可读存储介质
WO2022236877A1 (zh) * 2021-05-14 2022-11-17 广州广电运通金融电子股份有限公司 一种基于深度学习的指纹纹理提取方法、系统、装置及存储介质
CN115294141B (zh) * 2022-10-10 2023-03-10 惠智赋能(滨州)信息科技服务有限公司 一种基于声纳图像的深海渔网检测方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TW200715200A (en) * 2005-10-07 2007-04-16 Nanogate Optoelectronic Robot Inc System for detecting faked fingerprint
CN1991860A (zh) * 2005-12-26 2007-07-04 联杰光电股份有限公司 具有侦测毛细孔的指纹辨识系统及其方法
US20070177774A1 (en) * 2006-01-30 2007-08-02 Nanogate Optoelectronics Robot, Inc. Fingerprint Recognition System Having Detecting Pore and Method Thereof
CN105740750A (zh) * 2014-12-11 2016-07-06 深圳印象认知技术有限公司 指纹活体检测及识别方法与装置
CN106250890A (zh) * 2016-09-23 2016-12-21 南昌欧菲生物识别技术有限公司 一种指纹识别方法及装置
CN106682473A (zh) * 2016-12-20 2017-05-17 深圳芯启航科技有限公司 一种用户身份信息的识别方法及其装置

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8098906B2 (en) * 2006-10-10 2012-01-17 West Virginia University Research Corp., Wvu Office Of Technology Transfer & Wvu Business Incubator Regional fingerprint liveness detection systems and methods
CN100573553C (zh) * 2007-01-18 2009-12-23 中国科学院自动化研究所 基于薄板样条形变模型的活体指纹检测方法
CN100498822C (zh) * 2007-07-10 2009-06-10 深圳市天识科技有限公司 一种兼容不同指纹传感器图像信息的指纹识别方法
US8965069B2 (en) * 2011-09-30 2015-02-24 University Of Louisville Research Foundation, Inc. Three dimensional minutiae extraction in three dimensional scans
CN104463129B (zh) * 2014-12-17 2018-03-02 浙江维尔科技股份有限公司 一种指纹注册方法及装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TW200715200A (en) * 2005-10-07 2007-04-16 Nanogate Optoelectronic Robot Inc System for detecting faked fingerprint
CN1991860A (zh) * 2005-12-26 2007-07-04 联杰光电股份有限公司 具有侦测毛细孔的指纹辨识系统及其方法
US20070177774A1 (en) * 2006-01-30 2007-08-02 Nanogate Optoelectronics Robot, Inc. Fingerprint Recognition System Having Detecting Pore and Method Thereof
CN105740750A (zh) * 2014-12-11 2016-07-06 深圳印象认知技术有限公司 指纹活体检测及识别方法与装置
CN106250890A (zh) * 2016-09-23 2016-12-21 南昌欧菲生物识别技术有限公司 一种指纹识别方法及装置
CN106682473A (zh) * 2016-12-20 2017-05-17 深圳芯启航科技有限公司 一种用户身份信息的识别方法及其装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118171261A (zh) * 2023-10-09 2024-06-11 深圳市氿博传感科技有限公司 安防密码验证方法、装置、设备、介质及其计算机产品

Also Published As

Publication number Publication date
CN109840458A (zh) 2019-06-04

Similar Documents

Publication Publication Date Title
WO2019104932A1 (zh) 一种指纹识别方法及指纹采集设备
US10108858B2 (en) Texture features for biometric authentication
WO2021129323A1 (zh) 超声图像病灶描述方法、装置、计算机设备和存储介质
Liu et al. 3D fingerprint reconstruction system using feature correspondences and prior estimated finger model
US9390327B2 (en) Feature extraction and matching for biometric authentication
KR101985689B1 (ko) 선거리 계산 방법 및 장치
KR101632912B1 (ko) 지문 인식을 이용한 사용자 인증 방법
WO2016127736A1 (zh) 指纹重叠区域面积的计算方法及电子装置
TWI592881B (zh) 指紋辨識方法
BR122018007961B1 (pt) Método implementado por computador e sistema
WO2016018987A1 (en) Detecting specified image identifiers on objects
JP2009523266A (ja) 目に関して取得される画像中の虹彩を表す画素を特定する方法
JP2007188504A (ja) 画像中の画素強度をフィルタリングする方法
CN110378351B (zh) 印章鉴别方法及装置
Purnapatra et al. Presentation attack detection with advanced cnn models for noncontact-based fingerprint systems
CN112801031A (zh) 静脉图像识别方法、装置、电子设备及可读存储介质
KR20070088982A (ko) 변형에 강건한 홍채 인식 방법
Xu et al. Line segment detection with hough transform based on minimum entropy
Joshi et al. A novel approach implementation of eyelid detection in biometric applications
Babatunde et al. A block processing approach to fingerprint ridge-orientation estimation
JP5305366B2 (ja) 画像特徴抽出装置、画像特徴抽出方法、画像認識装置、及び画像認識方法
Heidari et al. Detection of Ridge Discontinuities in Fingerprint RecognitionInfluenced by Skin Diseases
CN118898517A (zh) 一种基于区块链和大数据的金融交易系统
KR101574409B1 (ko) 출판물의 핑거프린트 생성 방법과 이를 이용한 출판물 식별 장치 및 방법
Lodin et al. Automatic Iris Location Using Hough Transform.

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18884564

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18884564

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS (EPO FORM 1205A DATED 20.10.2020)

122 Ep: pct application non-entry in european phase

Ref document number: 18884564

Country of ref document: EP

Kind code of ref document: A1