WO2020187098A1 - 一种指纹图像增强、指纹识别和应用程序启动方法 - Google Patents

一种指纹图像增强、指纹识别和应用程序启动方法 Download PDF

Info

Publication number
WO2020187098A1
WO2020187098A1 PCT/CN2020/078648 CN2020078648W WO2020187098A1 WO 2020187098 A1 WO2020187098 A1 WO 2020187098A1 CN 2020078648 W CN2020078648 W CN 2020078648W WO 2020187098 A1 WO2020187098 A1 WO 2020187098A1
Authority
WO
WIPO (PCT)
Prior art keywords
fingerprint
image
fingerprint image
present application
current frame
Prior art date
Application number
PCT/CN2020/078648
Other languages
English (en)
French (fr)
Inventor
朱明铭
雷华
赵家财
梅丽
王进
Original Assignee
虹软科技股份有限公司
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 虹软科技股份有限公司 filed Critical 虹软科技股份有限公司
Priority to EP20774228.9A priority Critical patent/EP3940583A4/en
Priority to KR1020217033357A priority patent/KR20210136127A/ko
Priority to US17/041,465 priority patent/US11874907B2/en
Priority to JP2021555011A priority patent/JP7377879B2/ja
Publication of WO2020187098A1 publication Critical patent/WO2020187098A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/12—Fingerprints or palmprints
    • G06V40/13—Sensors therefor
    • G06V40/1318—Sensors therefor using electro-optical elements or layers, e.g. electroluminescent sensing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/31—User authentication
    • G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/60—Protecting data
    • G06F21/604—Tools and structures for managing or administering access control systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/60—Protecting data
    • G06F21/62—Protecting access to data via a platform, e.g. using keys or access control rules
    • G06F21/6209—Protecting access to data via a platform, e.g. using keys or access control rules to a single file or object, e.g. in a secure envelope, encrypted and accessed using a key, or with access control rules appended to the object itself
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/70—Denoising; Smoothing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/80—Geometric correction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/90—Dynamic range modification of images or parts thereof
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/30—Noise filtering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/54—Extraction of image or video features relating to texture
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/12—Fingerprints or palmprints
    • G06V40/1347—Preprocessing; Feature extraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • 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; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/12—Fingerprints or palmprints
    • G06V40/1365—Matching; Classification
    • G06V40/1376—Matching features related to ridge properties or fingerprint texture
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00—Network architectures or network communication protocols for network security
    • H04L63/08—Network architectures or network communication protocols for network security for authentication of entities
    • H04L63/0861—Network architectures or network communication protocols for network security for authentication of entities using biometrical features, e.g. fingerprint, retina-scan
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20024—Filtering details
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30196—Human being; Person

Definitions

  • the embodiments of the present application relate to image processing and terminal application technologies, in particular to a method for fingerprint image enhancement, fingerprint recognition, and application startup.
  • the under-screen fingerprint unlocking solution has the advantages of beautiful, convenient, fast unlocking speed, and conforming to user habits, and has become one of the mainstream ways of unlocking mobile phones.
  • the current under-screen fingerprint unlocking solutions due to unclear, incomplete or deformed fingerprint images, users need to repeatedly place their fingers in the fingerprint collection area during registration and collect multiple local fingerprints. It is more cumbersome, and because of the quality of the fingerprint image when unlocking, the unlocking efficiency is also low, which seriously affects the user experience.
  • the embodiments of the present application provide a method for fingerprint image enhancement, fingerprint identification, and application startup, which can reduce the number of fingerprint collections, reduce the cumbersomeness of fingerprint collection, improve the quality of fingerprint images, improve fingerprint unlocking efficiency, and improve user experience.
  • the embodiments of the present application provide a fingerprint image enhancement method, and the method may include:
  • the second preprocessing is performed on the effective area corrected by the direction field to obtain a fingerprint enhanced image.
  • the background texture is the average pixel value of N frames of fingerprint images before the current frame of fingerprint image, where N is a positive integer.
  • the removing the background texture of the fingerprint image to obtain a pure fingerprint image includes:
  • the first preprocessing includes performing contrast enhancement and/or denoising on the pure fingerprint image.
  • the acquisition of the effective area of the first preprocessed image adopts a preset fingerprint foreground segmentation algorithm.
  • the second preprocessing includes:
  • the fingerprint ridges in the binary image are refined to obtain the fingerprint enhanced image.
  • the embodiment of the present application also provides a fingerprint identification method, and the method may include:
  • the enhancement processing includes: eliminating the background texture of the fingerprint image of the current frame to obtain a pure fingerprint image;
  • the fingerprint identification is completed according to the comparison between the characteristic data and the characteristic data of the fingerprint template.
  • the enhancement processing may further include:
  • the second preprocessing is performed on the effective area corrected by the direction field to obtain a fingerprint enhanced image.
  • the background texture is the average pixel value of N frames of fingerprint images before the current frame of fingerprint image, where N is a positive integer.
  • the removing the background texture of the fingerprint image to obtain a pure fingerprint image may include:
  • the method may further include:
  • performing fingerprint distortion detection on the fingerprint enhanced image may include:
  • the fingerprint enhanced image is input into a classifier for classification, and the classification result includes a normal fingerprint image and a distorted fingerprint image.
  • performing distortion correction on the distorted fingerprint image may include:
  • the feature data may include minutiae feature and/or ridge feature of the fingerprint ridge.
  • completing fingerprint identification according to the comparison of the characteristic data and the characteristic data of the fingerprint template may include:
  • fingerprint recognition is completed.
  • the embodiment of the present application also provides a method for starting an application program based on fingerprint recognition, and the method for starting an application program may include:
  • the application startup method may further include: detecting whether the touch operation of the finger on the touch screen satisfies a preset condition for starting the fingerprint image collection step.
  • the method for starting an application program based on fingerprint recognition may further include: completing identity verification while starting the application program.
  • the embodiment of the present application also provides a fingerprint sensing system, which may include:
  • the display screen includes a light-emitting display unit for displaying pictures
  • the fingerprint collection module is set at least in a partial area below the display screen for collecting fingerprint images
  • the fingerprint identification module is configured to receive the fingerprint image and use the fingerprint identification method of any one of claims 8 to 17 to perform fingerprint identification on the fingerprint image.
  • the fingerprint collection module may include:
  • the imaging unit is arranged under the lens and used to directly obtain the fingerprint image on the display screen.
  • the fingerprint collection module is used to obtain a fingerprint image by detecting light emitted from the display screen and reflected on the surface of the finger back to the display screen.
  • the fingerprint collection module obtains a fingerprint image by detecting light that penetrates the display screen from a finger; wherein, when a light whose refraction angle is greater than a first threshold is detected, it is determined as the fingerprint ridge When a light with a refraction angle less than or equal to the first threshold is detected, it is determined to be a fingerprint valley line, and the fingerprint image is obtained according to the fingerprint ridge line and the fingerprint valley line.
  • the first threshold may be the refraction angle at the valley line of the fingerprint.
  • the fingerprint collection module may further include an optical path guide module, which may be used to guide light with a refraction angle greater than the first threshold.
  • the fingerprint collection module may further include a photoelectric sensor, which may be used to determine that the light is a fingerprint ridge when the light with a refraction angle greater than the first threshold is detected. When a light whose refraction angle is less than or equal to the first threshold is detected, it is determined that the light is a fingerprint valley line, thereby obtaining a fingerprint pattern.
  • the embodiment of the application also proposes an electronic device, which may include:
  • a memory for storing executable instructions of the processor
  • the processor is configured to execute the fingerprint identification method of any one of the foregoing by executing the executable instruction.
  • the electronic device may further include the fingerprint sensing system described in any one of the above.
  • the embodiment of the present application also proposes a storage medium, the storage medium may include a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute any one of the fingerprint identification methods described above.
  • the number of fingerprint collections is reduced, the cumbersomeness of fingerprint collection is reduced, the quality of fingerprint images is improved, the efficiency of fingerprint unlocking is improved, and the user experience is improved. Therefore, at least the following beneficial effects are included:
  • Fingerprint recognition can be performed at any position on the mobile phone screen, which is flexible and free.
  • Fig. 1 is a flowchart of a fingerprint image enhancement method according to an embodiment of the application
  • FIG. 2 is a flow chart of a method for correcting the wrong part in the initial direction field of the effective area by using the method of the direction field dictionary in an embodiment of the application;
  • FIG. 3 is a flowchart of a second preprocessing method according to an embodiment of the application.
  • FIG. 4 is a flowchart of a fingerprint identification method according to an embodiment of the application.
  • Fig. 5 is a flowchart of an enhancement processing method after performing background texture elimination on the fingerprint image of the current frame to obtain a pure fingerprint image according to an embodiment of the application;
  • FIG. 6 is a flow chart of a specific method for performing second preprocessing on the effective area corrected by the direction field according to an embodiment of the application;
  • FIG. 7 is a flowchart of a method before feature extraction on a fingerprint enhanced image according to an embodiment of the application.
  • FIG. 8 is a flowchart of a method for performing distortion correction on the distorted fingerprint image according to an embodiment of the application
  • FIG. 9-a is a first schematic diagram of ridges, minutiae points, and sub-ridges in an embodiment of the application
  • FIG. 9-b is a second schematic diagram of ridge lines, minutiae points, and sub-ridge lines according to an embodiment of the application.
  • FIG. 9-c is a schematic diagram of the label relationship of the sub-ridges in an embodiment of the application.
  • FIG. 10 is a flowchart of a method for starting an application program based on fingerprint recognition according to an embodiment of the application
  • FIG. 11 is a structural block diagram of a fingerprint sensing system according to an embodiment of the application.
  • FIG. 12 is a schematic structural diagram of the fingerprint sensing system of an embodiment of the application including a touch panel;
  • FIG. 13 is a schematic diagram of a method for obtaining a fingerprint image by detecting light emitted from the display screen and reflected on the surface of the finger back to the display screen according to an embodiment of the application;
  • FIG. 14 is a schematic diagram of a method for obtaining a fingerprint image by detecting light penetrating into a display screen from a finger according to an embodiment of the application;
  • FIG. 15 is a block diagram of the structure of an electronic device according to an embodiment of the application.
  • the embodiments of the present application provide a fingerprint image enhancement method. As shown in FIG. 1, the method may include S101-S105:
  • the solution of the embodiment of the present application may first perform enhancement processing on the collected fingerprint image.
  • the enhancement processing may include: performing background texture elimination on the fingerprint image of the current frame to obtain a pure fingerprint image.
  • the generally collected image not only contains fingerprints, but also contains the background texture of the fingerprint image (for example, the texture of the screen itself, the residual fingerprint image, etc.), so the enhancement processing can be performed first. Separate fingerprints from background texture.
  • the removing the background texture of the fingerprint image to obtain a pure fingerprint image may include:
  • the background texture may be characterized by the average value of pixels of N frames of fingerprint images before the fingerprint image of the current frame, where N is a positive integer.
  • the method of averaging the pixels of multiple frames of images can be used to obtain the background texture, that is, the pixel average is calculated using the sequence of N frames of fingerprint images before the fingerprint image of the current frame, and the result obtained can be approximated as Background texture.
  • the background texture is relatively fixed and the intensity is relatively strong, while the fingerprint changes greatly and the intensity is weak.
  • the averaging of multiple frames can further weaken the accidental fingerprints and retain a relatively stable background texture.
  • the current frame fingerprint image and the background texture can be used to perform pixel corresponding subtraction to obtain a pure fingerprint image after removing the background texture.
  • performing pixel corresponding subtraction on the background texture in the fingerprint image of the current frame may include: subtracting the pixels of the fingerprint image of the current frame from the pixels of the previous N frames of the fingerprint image of the current frame The average value is obtained, and then the background texture in the fingerprint image of the current frame is reduced according to a specific algorithm (for example, multiplying by a preset coefficient and adding to a preset value).
  • the fingerprint image enhancement method may further include: before eliminating the background texture of the current frame fingerprint image, performing local color transfer on the current frame fingerprint image and the background texture to Keep the brightness of the current frame fingerprint image and the background texture consistent.
  • the possible problem of performing background texture elimination on the fingerprint image of the current frame is that the brightness of the fingerprint image of the current frame is inconsistent with the background texture, and direct subtraction may produce wrong results. Therefore, the fingerprint image and the background texture can be transferred locally to keep the brightness of the background texture consistent with the fingerprint image, and then corresponding pixel-by-pixel reduction can be performed to obtain a relatively pure fingerprint image, that is, the above-mentioned pure fingerprint image.
  • the first preprocessing may include performing contrast enhancement and/or denoising on the pure fingerprint image.
  • performing contrast enhancement and denoising on the pure fingerprint image may include:
  • a preset denoising algorithm is used to perform the first filtering on the pure fingerprint image after the contrast is enhanced.
  • the fingerprint image obtained after the background texture is reduced usually has a low contrast, and sometimes the overall contrast is uneven. Therefore, local contrast normalization (LCN) or local adaptive histogram equalization processing can be applied to the pure fingerprint image after the background texture is reduced to enhance the contrast of the image and make the overall contrast of the image relatively uniform.
  • LN local contrast normalization
  • the preset denoising algorithm can be used to take additional denoising processing to suppress the noise of the fingerprint image after the contrast is enhanced. So far, a relatively clear first preprocessed image is obtained.
  • the preset denoising algorithm may include: Fast Non-Local Means Denoising (Fast Non-Local Means Denoising).
  • the quality of the obtained fingerprint image is already relatively good, but further processing can be done.
  • the effective area in the pure fingerprint image after the first filtering can be obtained; the effective area is obtained by calculating the pure fingerprint image according to a preset fingerprint foreground segmentation algorithm .
  • the effective area in a pure fingerprint image refers to the middle part of the fingerprint image, and the surrounding area of the fingerprint image is often an invalid background part. Processing the invalid part will not only increase time overhead, but also It may cause additional interference. Therefore, the fingerprint image can be segmented, the effective fingerprint foreground can be extracted, and the invalid background part can be removed.
  • the acquisition of the effective area of the first preprocessed image may adopt a preset fingerprint foreground segmentation algorithm.
  • the preset fingerprint foreground segmentation algorithm may include: an improved gradient-based fingerprint image segmentation algorithm (Fingerprint Image Segmentation Based on Boundary Values), or based on the average gray value of the image block Or the segmentation method of gray variance.
  • the preset fingerprint foreground segmentation algorithm can improve the performance of subsequent processing while eliminating unnecessary interference. After obtaining the fingerprint foreground area (that is, the above-mentioned effective area), normalizing the area is performed to remove the difference in image intensity caused by different pressing forces during fingerprint collection.
  • S104 Perform direction field estimation and direction field correction on the effective area.
  • the direction field estimation can be performed on the effective area in the pure fingerprint image.
  • the direction field is an inherent attribute of the fingerprint image, which defines the invariant coordinates of the ridge and valley of the fingerprint in the local neighborhood.
  • the direction field estimation may use Fourier transform, gradient method, etc. to estimate the initial direction field of the fingerprint image; the direction field correction may include: Correct the wrong part in the initial direction field of the effective area to obtain more accurate direction field results.
  • the method of adopting the direction field dictionary to correct the erroneous part in the initial direction field of the effective area may include S201-S203:
  • the preset direction field dictionary is obtained by training sample fingerprint images meeting preset quality requirements, extracting the direction distribution characteristics of all blocks in the sample fingerprint image and performing clustering.
  • the fingerprint image can be divided into multiple blocks, and the direction distribution characteristics in each block can be calculated, so that each block is called a word in the direction field dictionary.
  • the fingerprint image with better quality can be used for training in advance, and the direction distribution characteristics in all blocks can be extracted and clustered to obtain a complete direction field dictionary.
  • the direction distribution of each block in the dictionary is relatively continuous and smooth.
  • the blocks are divided in the same way and compared with the direction field dictionary obtained by training.
  • the most similar block in the direction field dictionary is used to correct the direction distribution of the current block to make the current fingerprint
  • the direction field in the image is more accurate.
  • the second preprocessing may include S301-S303:
  • a Gabor filter may be used to perform the second filtering on the effective area corrected by the direction field to remove noise in the image and preserve the sinusoidal ridges and valleys.
  • an adaptive image binarization algorithm may be used to calculate optimal thresholds for different regions in the effective region after the second filtering, to obtain a complete binary image of the fingerprint information in the effective region. Valued image.
  • the ridge line in the binarized image can be thinned to the width of one pixel, and the original topological structure of the fingerprint is retained without adding additional noise.
  • a Gabor filter can be used to filter the fingerprint foreground, which can remove part of the noise in the image and retain sinusoidal ridges and valleys. Then perform the binarization operation on the filtered fingerprint image. During this period, it is very important to select the appropriate threshold. You can use the adaptive image binarization algorithm to calculate the optimal threshold for different regions to obtain a binarized image with complete fingerprint information. .
  • the last step of fingerprint image preprocessing can be refinement. This step can refine the ridges in the binary image to the width of one pixel while preserving the original topological structure of the fingerprint without adding additional noise to facilitate subsequent steps. Feature extraction.
  • the embodiments of the present application provide a fingerprint identification method, as shown in FIG. 4, the method may include S401-S403:
  • S401 Perform enhancement processing on the captured fingerprint image of the current frame; the enhancement processing includes: performing background texture elimination on the fingerprint image of the current frame to obtain a pure fingerprint image.
  • the solution of the embodiment of the present application may first perform enhancement processing on the collected fingerprint image.
  • the enhancement processing may include: performing background texture elimination on the fingerprint image of the current frame to obtain a pure fingerprint image.
  • the generally collected image not only contains fingerprints, but also contains the background texture of the fingerprint image (for example, the texture of the screen itself, the residual fingerprint image, etc.), so the enhancement processing can be performed first. Separate fingerprints from background texture.
  • the performing background texture elimination on the fingerprint image of the current frame to obtain a pure fingerprint image may include:
  • N is a positive integer
  • the method of averaging the pixels of multiple frames of images can be used to obtain the background texture, that is, the pixel average is calculated using the sequence of N frames of fingerprint images before the fingerprint image of the current frame, and the result obtained can be approximated as Background texture.
  • the background texture is relatively fixed and the intensity is relatively strong, while the fingerprint changes greatly and the intensity is weak.
  • the averaging of multiple frames can further weaken the accidental fingerprints and retain a relatively stable background texture.
  • the current frame fingerprint image and the background texture can be used to perform pixel corresponding subtraction to obtain a pure fingerprint image after removing the background texture.
  • performing pixel corresponding subtraction on the background texture in the fingerprint image of the current frame may include: subtracting the pixels of the fingerprint image of the current frame from the pixels of the previous N frames of the fingerprint image of the current frame The average value is obtained, and then the background texture in the fingerprint image of the current frame is reduced according to a specific algorithm (for example, multiplying by a preset coefficient and adding to a preset value).
  • the method may further include:
  • the possible problem of performing background texture elimination on the fingerprint image of the current frame is that the brightness of the fingerprint image of the current frame is inconsistent with the background texture, and direct subtraction may produce wrong results. Therefore, the fingerprint image and the background texture can be transferred locally to keep the brightness of the background texture consistent with the fingerprint image, and then corresponding pixel-by-pixel reduction can be performed to obtain a relatively pure fingerprint image, that is, the above-mentioned pure fingerprint image.
  • the enhancement processing may further include S501-S504:
  • S501 Perform first preprocessing on the pure fingerprint image to obtain a first preprocessed image.
  • the first preprocessing may include:
  • a preset denoising algorithm is used to perform the first filtering on the pure fingerprint image after the contrast is enhanced.
  • the fingerprint image obtained after the background texture is reduced usually has a low contrast, and sometimes the overall contrast is uneven. Therefore, local contrast normalization (LCN) or local adaptive histogram equalization processing can be applied to the pure fingerprint image after the background texture is reduced to enhance the contrast of the image and make the overall contrast of the image relatively uniform.
  • LN local contrast normalization
  • the preset denoising algorithm can be used to take additional denoising processing to suppress the noise of the fingerprint image after the contrast is enhanced. So far, a relatively clear first preprocessed image is obtained.
  • the preset denoising algorithm may include: Fast Non-Local Means Denoising (Fast Non-Local Means Denoising).
  • the quality of the fingerprint image obtained is relatively good, but further processing can be performed, for example, obtaining the pure fingerprint after the first filtering.
  • the effective area in the image; the effective area is obtained by calculating the pure fingerprint image according to a preset fingerprint foreground segmentation algorithm.
  • the effective area in a pure fingerprint image refers to the middle part of the fingerprint image, and the surrounding area of the fingerprint image is often an invalid background part. Processing the invalid part will not only increase time overhead, but also It may cause additional interference. Therefore, the fingerprint image can be segmented, the effective fingerprint foreground can be extracted, and the invalid background part can be removed.
  • the preset fingerprint foreground segmentation algorithm may include: an improved gradient-based fingerprint image segmentation algorithm (Fingerprint Image Segmentation Based on Boundary Values), or based on the average gray value of the image block Or the segmentation method of gray variance.
  • the preset fingerprint foreground segmentation algorithm can improve the performance of subsequent processing while eliminating unnecessary interference. After obtaining the fingerprint foreground area (that is, the above-mentioned effective area), normalizing the area is performed to remove the difference in image intensity caused by different pressing forces during fingerprint collection.
  • S503 Perform direction field estimation and direction field correction on the effective area.
  • the direction field estimation can be performed on the effective area in the pure fingerprint image.
  • the direction field is an inherent attribute of the fingerprint image, which defines the invariant coordinates of the ridge and valley of the fingerprint in the local neighborhood.
  • the direction field estimation may use Fourier transform, gradient method, etc. to estimate the initial direction field of the fingerprint image; the direction field correction may include: Correct the wrong part in the initial direction field of the effective area to obtain more accurate direction field results.
  • the method of adopting a direction field dictionary to correct an erroneous part in the initial direction field of the effective area may include:
  • the preset direction field dictionary is obtained by training sample fingerprint images that meet preset quality requirements, extracting the directional distribution characteristics of all blocks in the sample fingerprint image and performing clustering.
  • the fingerprint image can be divided into multiple blocks, and the direction distribution characteristics in each block can be calculated, so that each block is called a word in the direction field dictionary.
  • the fingerprint image with better quality can be used for training in advance, and the direction distribution characteristics in all blocks can be extracted and clustered to obtain a complete direction field dictionary.
  • the direction distribution of each block in the dictionary is relatively continuous and smooth.
  • the blocks are divided in the same way and compared with the direction field dictionary obtained by training.
  • the most similar block in the direction field dictionary is used to correct the direction distribution of the current block to make the current fingerprint
  • the direction field in the image is more accurate.
  • S504 Perform a second preprocessing on the effective region corrected by the direction field to obtain a fingerprint enhanced image.
  • the second preprocessing may include: second filtering, binarization processing, and refinement.
  • the performing the second preprocessing on the effective area corrected by the direction field may specifically include S601-S603:
  • S601 Use a Gabor filter to perform the second filtering on the effective area corrected by the direction field, so as to remove noise in the image and retain sinusoidal ridges and valleys;
  • S603 Refine the ridge line in the binarized image to the width of one pixel, and retain the original topological structure of the fingerprint without adding additional noise.
  • a Gabor filter can be used to filter the fingerprint foreground, which can remove part of the noise in the image and retain sinusoidal ridges and valleys. Then perform the binarization operation on the filtered fingerprint image. During this period, it is very important to select the appropriate threshold. You can use the adaptive image binarization algorithm to calculate the optimal threshold for different regions to obtain a binarized image with complete fingerprint information. .
  • the last step of fingerprint image preprocessing can be refinement. This step can refine the ridges in the binary image to the width of one pixel while preserving the original topological structure of the fingerprint without adding additional noise to facilitate subsequent steps. Feature extraction.
  • S701-S702 may be further included:
  • S702 Perform distortion correction on the distorted fingerprint image.
  • the problem of fingerprint distortion and deformation is usually encountered during the fingerprint identification process. This is caused by the different force and direction of the finger pressing during the collection process, which will cause the same finger to produce different Feature data affect the final recognition result. Therefore, fingerprint distortion detection and correction algorithms can be used to correct the distorted fingerprint image to an undistorted state, thereby ensuring the consistency of the final feature data.
  • the performing fingerprint distortion detection on the fingerprint enhanced image may include:
  • the fingerprint enhanced image is input into a classifier for classification, and the classification result includes a normal fingerprint image and a distorted fingerprint image.
  • the distortion of the fingerprint may cause the final extracted feature data to be different from the normal state, which greatly reduces the matching score and causes an incorrect recognition result. Therefore, the fingerprint image can be detected for distortion first, and if the distortion is detected, the fingerprint image is corrected to restore the fingerprint image to a normal state.
  • a large number of normal fingerprint images and distorted fingerprint images collected in advance can be used to train a classifier, and the enhanced fingerprint image can be input into the trained classifier, and the currently input fingerprint image can be divided into two categories. If the classification result is distorted, distortion correction is performed on the fingerprint image.
  • the performing distortion correction on the distorted fingerprint image may include S801-S803:
  • S803 Perform inverse transformation correction on the twisted fingerprint image according to the reference twisted fingerprint.
  • the distortion correction may be completed by estimating the distortion field of the distorted fingerprint image and inversely transforming it.
  • a database set which contains the distortion field corresponding to various distortion fingerprints (distortion field refers to the transformation relationship between a fingerprint from a normal undistorted state to a distorted state), direction field and periodogram (period The graph refers to the ridge period or frequency (representing the density of ridges) at each position in the fingerprint image.
  • Specific methods can include: collecting common image pairs of normal fingerprints and distorted fingerprints, so as to obtain statistical models of common distorted fields through these image pairs, and use these statistical models to synthesize a large number of distorted fields and act on the normal fingerprint images to obtain normal fingerprints.
  • the set of twisted fingerprint images corresponding to the images and their direction fields and periodograms is used as the aforementioned database set.
  • the distortion correction of the fingerprint image may include: for the currently detected distortion fingerprint image, its direction field and periodogram may be extracted first, and then the characteristics of the current distortion fingerprint image may be searched in the database. According to the closest reference twisted fingerprint, the current twisted fingerprint image is inversely transformed and corrected according to the twisted field corresponding to the twisted fingerprint to restore the current twisted fingerprint image to a normal state.
  • S402 Perform feature extraction on the fingerprint enhanced image to obtain feature data.
  • the feature data may include, but is not limited to: minutiae point features and ridge line features of fingerprint ridges.
  • the fingerprint image enhancement processing can be completed after the above steps, and the distortion of the fingerprint can be corrected, thereby obtaining a high-quality fingerprint image.
  • features can be extracted from the refined fingerprint ridges to obtain feature data.
  • minutiae features may be extracted, and the minutiae features may include end points and bifurcation points of ridges and the like.
  • the feature data of the current fingerprint image can be obtained.
  • S403 Perform fingerprint identification according to the comparison between the characteristic data and the characteristic data of the fingerprint template.
  • the completion of fingerprint identification based on the comparison of the characteristic data with the characteristic data of the fingerprint template may include:
  • fingerprint recognition is completed.
  • the similarity between different fingerprints can be calculated, and the similarity can be calculated according to the similarity.
  • the internally stored fingerprint template that is, the preset feature data of different fingerprints
  • the internally stored fingerprint template that is, the preset feature data of different fingerprints
  • the adopted sub-structure can be minutiae points and multiple related ridge lines, specifically: minutiae points, minutiae points A substructure formed by the ridge line and the ridge lines adjacent to both sides of the ridge line.
  • the minutia points (the end points or bifurcation points inside the ridge line represent minutiae points, as shown by the black dots in Figure 9-a and 9-b), and then draw a line along the vertical ridge line through the minutia points
  • the two intersection points of a straight line, a straight line and adjacent ridges are called projection points.
  • the ridges in the substructure are split through the minutiae points and projection points, and they are labeled according to their relative positions and directions, as shown in Figure 9-a , As shown in Figure 9-b.
  • the original complete black line in the figure represents the ridge line (2+3, 4+5, 1 in Figure 9-a and 4+5, 3+1, 3+2, 6+7 in Figure 9-b),
  • the end points or bifurcation points inside the ridge line represent the detail points (shown as the black dots in Figure 9-a and 9-b).
  • the white points in the figure are the projection points.
  • the ridge line is split by the detail points or projection points to obtain sub-ridges Lines (as shown in 2, 3, 1, 4, 5 in 9-a and 4, 5, 3, 1, 2, 6, 7 in Figure 9-b).
  • the originally matched pair of minutiae points and sub-structure pairs should be completely overlapped after the alignment transformation, but in fact, due to the error of the minutiae extraction process and the alignment transformation and the real physics The error of the transformation results in the incomplete matching between the pair of minutiae points and the pair of substructures. Therefore, a more stable matching scheme can be used to calculate the similarity between fingerprints.
  • the following two aspects can be mainly considered: 1.
  • Minutiae pair aspect You can first select a reference minutiae point and convert all other minutiae points into The reference minutiae point is the polar coordinate of the origin; then all minutiae points can be connected in an ascending order of angle to form a feature string; finally, the edit distance between the fingerprint template feature string and the current fingerprint feature string can be calculated, according to the edit distance Determine the matching score between pairs of minutiae points.
  • Sub-structure pair aspect It can traverse the first N pairs of sub-structures that are most matched when fingerprints are aligned.
  • the corresponding ridges in each pair of sub-structures form the initial matching ridge pair, and the matching ridge pairs are adjacent
  • the ridge pair constitutes a new matching ridge pair, thus obtaining the matching ridge pair set of the two fingerprints.
  • the matching score between the substructure pairs can be obtained by the proportion of matching points on the matching ridge pair and the proportion of matching minutiae points in the matching substructure pairs.
  • the minutiae pair matching score and the substructure pair matching score can be integrated to obtain the final similarity between two fingerprint images. By comparing the seeded similarity with the preset similarity threshold, the current two can be confirmed Whether the fingerprint is matched successfully.
  • a method for starting an application program based on fingerprint recognition is also provided.
  • the method for starting an application program may further include S1001-S1003:
  • the application startup method may further include, before the fingerprint image is collected, detecting whether the touch operation of the finger on the touch screen satisfies a preset condition for initiating the fingerprint image collection step.
  • the application program may be a computer program that allows only authorized personnel to access to protect user privacy, personal information, or confidential information of enterprises and institutions.
  • the application startup method can also complete identity verification while starting the application.
  • the user identity information corresponding to the application and the fingerprint can be collected when the application is successfully opened by fingerprint recognition, and used for big data analysis.
  • the user's preferences and habits can be analyzed by collecting information such as the frequency and time of use of the application by the user, which can help application developers to make market planning.
  • the application is first opened by tapping or touching with a finger, and then when the application prompts for fingerprint verification, pressing the finger on the specific fingerprint recognition area again can complete the identity verification. It can be seen that in this traditional method, at least two steps are required to start the application and complete the identity verification. The operation is cumbersome and time-consuming, which reduces the user experience to a certain extent.
  • the identity verification can be completed at the same time as the application is started, so as to confirm that the user has the authority to perform corresponding operations on the application and provide access to the application. Secure access to programs (for example, secure financial transactions).
  • a fingerprint sensing system 1 is also provided. As shown in FIG. 11, the fingerprint sensing system may include: a display screen 11, a fingerprint acquisition module 12, and a fingerprint recognition module 13. .
  • the display screen 11 includes a light-emitting display unit for displaying pictures.
  • the light-emitting display unit may be a self-luminous display unit, such as a light-emitting diode (Light-Emitting Diode, LED), an organic light-emitting diode (Organic Light-Emitting Diode, OLED), or a micro light-emitting diode (Micro LED). -LED) etc.
  • the light-emitting display unit may also be a passive light-emitting display unit, such as a liquid crystal display (LCD) or the like.
  • the fingerprint collection module 12 is arranged at least in a partial area below the display screen for collecting fingerprint images.
  • the fingerprint collection module can be arranged at least in a partial area below the display screen to Reduce the occupation of the display area.
  • the fingerprint recognition module 13 is used to receive the fingerprint image of the current frame, and use the above fingerprint recognition method to perform fingerprint recognition on the fingerprint image.
  • the display screen 11 may be a touch display screen, which can not only perform screen display, but also detect the operation of the user's finger (such as touching, pressing, or approaching the display screen), thereby providing the user with a person Computer interactive interface.
  • the fingerprint sensing system may further include a touch panel (Touch Panel, TP).
  • the touch panel may be provided on the surface of the display screen, or may be partially or wholly integrated in The display screen constitutes a touch screen.
  • the fingerprint sensing system 1 may further include a cover plate, which is arranged above the display screen as an interface for the user to touch and display the screen, so as to protect the display screen.
  • the cover plate can be glass or sapphire, and is not limited thereto.
  • the fingerprint collection module 12 may include an optical collimator and a photodetector (Photo Detector). Through the optical collimator, only the light whose incident angle is smaller than the preset angle can reach the photodetector.
  • the fingerprint acquisition module may include a lens and an imaging unit, wherein the imaging unit may be arranged under the lens, and the lens imaging principle is used to directly obtain Fingerprint image on the display.
  • the lens can be a convex lens.
  • one or more lenses and imaging units can be respectively arranged under the display, so as to realize fingerprint collection and recognition in a partial area of the screen, half screen or full screen.
  • the lens and the imaging unit may be independent components or an integrated component. When the lens and the imaging unit are independent of each other, the numbers of the two may not necessarily correspond to one another.
  • the fingerprint acquisition module is used to obtain fingerprints by detecting light emitted from the display screen and reflected on the surface of the finger back to the display screen.
  • the fingerprint acquisition module may include a first photoelectric sensor, and the first photoelectric sensor may obtain a fingerprint image by detecting light emitted from the display screen and reflected on the surface of the finger back to the display screen.
  • the refractive index at the ridge line of the fingerprint is greater than the refractive index at the valley line of the fingerprint, so the fingerprint valley line will form Total reflection, there is no total reflection at the fingerprint ridge line, but a part of the light will be transmitted to the inside of the finger, which will cause the intensity of the reflected light at the fingerprint valley line to be greater than the reflected light at the fingerprint ridge line.
  • the intensity of the light reflected back to the display screen can determine the fingerprint ridgeline and fingerprint valley line to obtain a fingerprint image.
  • the excitation light is a built-in luminous display unit and the intensity of the reflected light needs to be detected, the ambient light has a greater impact on the effect.
  • the difference in the intensity of the reflected light is small, the fingerprint image will be reduced. Clarity.
  • the fingerprint collection module 12 can also obtain a fingerprint by detecting light that penetrates into the display screen from a finger. An image, wherein when a light with a refraction angle greater than the first threshold is detected, the light can be determined as a fingerprint ridge, otherwise the light is determined as a fingerprint valley, and the fingerprint image is obtained according to the fingerprint ridge and fingerprint valley .
  • the first threshold may be the refraction angle at the valley line of the fingerprint.
  • the fingerprint collection module 12 may include an optical path guiding module for guiding light with a refraction angle greater than the first threshold.
  • the fingerprint collection module 12 may also include a photoelectric sensor (a second photoelectric sensor).
  • the second photoelectric sensor can be used to determine the fingerprint ridge line when the light whose refraction angle is greater than the first threshold is detected, otherwise it is the fingerprint valley line. , Thereby obtaining fingerprint patterns.
  • the second photosensor may be a complementary metal oxide semiconductor CMOS sensor, a thin film transistor TFT sensor, or other customized sensors.
  • CMOS complementary metal oxide semiconductor
  • TFT thin film transistor
  • the excitation light itself is ambient light
  • the acquisition of the fingerprint image is not affected by ambient light.
  • the stronger the ambient light the better the collection effect.
  • the light-emitting display unit built into the display screen around the finger can be illuminated or the external light source can be illuminated to enhance the intensity of the light.
  • the display area of the display screen can be extended to the entire surface of the electronic device.
  • the fingerprint sensing system can be installed in a partial area or the entire area below the display screen, thereby realizing partial area, half-screen or full-screen fingerprint recognition.
  • the fingerprint sensing system is usually set in the area outside the display screen and the contact surface with the finger is small, such as the Apple iPhone 6, which has great restrictions on the identification object and identification method, and is not suitable for collecting and identifying large patterns. (For example, palm prints), it is not suitable for simultaneous identification and verification of multiple identification objects.
  • one or more independent or integral fingerprint sensing systems can be set in the partial area, half-screen area, or full-screen area under the display screen, it is possible to realize the detection of large patterns (for example, Palmprints) are collected and identified to expand the application scenarios; and multiple identification objects can be identified and verified at the same time to enhance the security of the application.
  • a financial payment application can be set to verify the fingerprints of two people at the same time. In order to start and complete the payment, you can set the full-screen fingerprint recognition, so that the fingers of two people can touch any area of the display at the same time.
  • the technical solutions of the embodiments of the present application can be applied to various electronic devices with display screens, such as smart phones, notebook computers, tablet computers, digital cameras, game consoles, smart bracelets, and smart phones.
  • Portable or wearable electronic devices such as watches, and other electronic devices such as automated teller machines (ATM), information management systems, electronic door locks, etc.
  • ATM automated teller machines
  • the technical solutions of the embodiments of the present application can also perform other biometric identifications other than fingerprints, which are not limited in the embodiments of the present application.
  • an electronic device A is also provided. As shown in FIG. 15, the electronic device may include:
  • the memory 3 is configured to store executable instructions of the processor 2;
  • the processor 2 is configured to execute the fingerprint identification method of any one of the foregoing by executing the executable instruction.
  • the electronic device A may further include the fingerprint sensing system 1 described in any one of the above.
  • the storage medium includes a stored program, wherein when the program runs, the device where the storage medium is located is controlled to execute any one of the foregoing contents.
  • Fingerprint recognition can be performed at any position on the mobile phone screen, which is flexible and free.
  • Such software may be distributed on a computer-readable medium
  • the computer-readable medium may include a computer storage medium (or non-transitory medium) and a communication medium (or transitory medium).
  • the term computer storage medium includes volatile and non-volatile memory implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data).
  • Computer storage media include but are not limited to RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassette, tape, magnetic disk storage or other magnetic storage device, or Any other medium used to store desired information and that can be accessed by a computer.
  • communication media usually contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as carrier waves or other transmission mechanisms, and may include any information delivery media .

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computer Security & Cryptography (AREA)
  • Human Computer Interaction (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Software Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Bioethics (AREA)
  • Computing Systems (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Automation & Control Theory (AREA)
  • Biomedical Technology (AREA)
  • Collating Specific Patterns (AREA)
  • Geometry (AREA)
  • Image Input (AREA)
  • Image Processing (AREA)
  • Image Analysis (AREA)

Abstract

一种指纹图像增强、指纹识别和应用程序启动方法，该指纹图像增强方法包括：消除当前帧指纹图像的背景纹理，获取纯指纹图像（S101）；对所述纯指纹图像进行第一预处理，获得第一预处理图像（S102）；获取所述第一预处理图像的有效区域（S103）；对所述有效区域进行方向场估计和方向场校正（S104）；对经过方向场校正的所述有效区域进行第二预处理，获得指纹增强图像（S105）。可有效减少了指纹采集次数，提高指纹图像的质量，降低指纹采集的繁琐度，提高指纹解锁效率，提高用户体验感。

Description

一种指纹图像增强、指纹识别和应用程序启动方法
本申请要求于2019年3月15日提交至中国国家知识产权局、申请号为201910198302.7、发明名称为“一种指纹图像增强、指纹识别和应用程序启动方法”的专利申请的优先权，其全部内容通过引用结合在本公开中。
技术领域
本申请实施例涉及图像处理与终端应用技术，尤指一种指纹图像增强、指纹识别和应用程序启动方法。
背景技术
伴随着移动终端(如手机)从功能机向智能机的发展转变，移动终端的解锁方式也在不断发生改变，从最早的以数字密码或图案为主的解锁方式，向以指纹、人脸、虹膜等为主的生物识别解锁方式转变。在技术不断突破的同时，解锁的便捷性与安全性也得到大幅提升。
如今，屏下指纹解锁方案拥有美观、便捷、解锁速度快、符合用户习惯等优点，成为当下手机解锁的主流方式之一。但是，目前的绝大多数屏下指纹解锁方案中，由于采集的指纹图像不清晰、不完整或变形等原因，使得用户在注册时需要反复将手指放置到指纹采集区域，采集多次局部指纹，较为繁琐，并且在解锁时由于指纹图像的质量问题也导致解锁效率低，严重影响用户体验感。
发明内容
本申请实施例提供了一种指纹图像增强、指纹识别和应用程序启动方法，能够减少指纹采集次数，降低指纹采集的繁琐度，提高指纹图像的质量，提高指纹解锁效率，提高用户体验感。
为了达到本申请实施例目的，本申请实施例提供了一种指纹图像增强方法，所述方法可以包括：
消除当前帧指纹图像的背景纹理，获取纯指纹图像；
对所述纯指纹图像进行第一预处理，获得第一预处理图像；
获取所述第一预处理图像的有效区域；
对所述有效区域进行方向场估计和方向场校正；
对经过方向场校正的所述有效区域进行第二预处理，获得指纹增强图像。
在本申请的示例性实施例中，所述背景纹理为所述当前帧指纹图像之前N帧指纹图像的像素均值，其中，N为正整数。
在本申请的示例性实施例中，，所述消除指纹图像的背景纹理，获得纯指纹图像包括：
在所述当前帧指纹图像中对所述背景纹理进行像素对应消减，获取所述纯指纹图像。
在本申请的示例性实施例中，在消除当前帧指纹图像的背景纹理之前，对所述当前帧指纹图像和所述背景纹理进行局部颜色传递。
在本申请的示例性实施例中，第一预处理包括对所述纯指纹图像进行对比度增强和/或去噪。
在本申请的示例性实施例中，所述获取所述第一预处理图像的有效区域采用预设的指纹前景分割算法。
在本申请的示例性实施例中，所述第二预处理包括：
对所述有效区域去噪；
获取二值化图像；
将所述二值化图像中的指纹脊线进行细化，获得所述指纹增强图像。
本申请实施例还提供了一种指纹识别方法，所述方法可以包括：
对采集的当前帧指纹图像进行增强处理，获得指纹增强图像；所述增强 处理包括：消除当前帧指纹图像的背景纹理，获得纯指纹图像；
对所述指纹增强图像进行特征提取，获取特征数据；
根据所述特征数据与指纹模板的特征数据的比对完成指纹识别。
在本申请的示例性实施例中，所述增强处理还可以包括：
对所述纯指纹图像进行第一预处理，获得第一预处理图像；
获取所述第一预处理图像的有效区域；
对所述有效区域进行方向场估计和方向场校正；
对经过方向场校正的所述有效区域进行第二预处理，获得指纹增强图像。
在本申请的示例性实施例中，所述背景纹理为所述当前帧指纹图像之前N帧指纹图像的像素均值，其中，N为正整数。
在本申请的示例性实施例中，所述消除指纹图像的背景纹理，获得纯指纹图像可以包括：
在所述当前帧指纹图像中对所述背景纹理进行像素对应消减，获取所述纯指纹图像。
在本申请的示例性实施例中，在消除当前帧指纹图像的背景纹理之前，对所述当前帧指纹图像和所述背景纹理进行局部颜色传递。
在本申请的示例性实施例中，在对所述指纹增强图像进行特征提取之前还可以包括：
对所述指纹增强图像进行指纹扭曲检测，确定所述指纹增强图像为正常指纹图像或扭曲指纹图像；
对所述扭曲指纹图像进行扭曲校正。
在本申请的示例性实施例中，对所述指纹增强图像进行指纹扭曲检测可以包括：
将所述指纹增强图像输入分类器进行分类，分类结果包括正常指纹图像和扭曲指纹图像。
在本申请的示例性实施例中，对所述扭曲指纹图像进行扭曲校正可以包括：
提取所述扭曲指纹图像的方向场和周期图；
根据所述方向场和周期图在数据库集中查找与所述扭曲指纹图像最接近的参考扭曲指纹；
根据所述参考扭曲指纹对所述扭曲指纹图像进行逆变换校正。
在本申请的示例性实施例中，所述特征数据可以包括指纹脊线的细节点特征和/或脊线特征。
在本申请的示例性实施例中，根据所述特征数据与指纹模板的特征数据的比对完成指纹识别可以包括：
计算所述特征数据与所述指纹模板的特征数据的特征相似度；
当所述特征相似度大于或等于特征相似度阈值时，完成指纹识别。
本申请实施例还提供了一种基于指纹识别的应用程序启动方法，所述应用程序启动方法可以包括：
采集指纹图像；
采用上述内容中任意一项所述的指纹识别方法，对所述指纹图像进行指纹识别；
当指纹识别成功时开启所述应用程序。
在本申请的示例性实施例中，在所述采集指纹图像之前，所述应用程序启动方法还可以包括：检测触控屏上手指的触控操作是否满足启动采集指纹图像步骤的预设条件。
在本申请的示例性实施例中，所述基于指纹识别的应用程序启动方法还可以包括：在开启所述应用程序的同时完成身份验证。
本申请实施例还提供了一种指纹感测系统，可以包括：
显示屏，包括发光显示单元，用于显示画面；
指纹采集模组，至少设置在所述显示屏下方的局部区域，用于采集指纹图像；
指纹识别模组，用于接收所述指纹图像，并采用权利要求8至17中任意一项所述的指纹识别方法，对所述指纹图像进行指纹识别。
在本申请的示例性实施例中，所述指纹采集模组可以包括：
透镜；
成像单元，设置于透镜下方，用于直接获取显示屏上的指纹图像。
在本申请的示例性实施例中，所述指纹采集模组用于通过检测自显示屏发出并在手指表面反射回显示屏的光线来获得指纹图像。
在本申请的示例性实施例中，所述指纹采集模组通过检测从手指透入显示屏的光线获得指纹图像；其中，当检测到折射角大于第一阈值的光线时，确定为指纹脊线，当检测到折射角小于或等于所述第一阈值的光线时，确定为指纹谷线，根据所述指纹脊线和所述指纹谷线获得所述指纹图像。
在本申请的示例性实施例中，所述第一阈值可以为所述指纹谷线处的折射角。
在本申请的示例性实施例中，所述指纹采集模组还可以包括光路引导模组，可以用于引导折射角大于所述第一阈值的光线。
在本申请的示例性实施例中，所述指纹采集模组还可以包括光电传感器，可以用于在检测到折射角大于所述第一阈值的光线时，确定所述光线为指纹脊线，在检测到折射角小于或等于所述第一阈值的光线时，确定所述光线为指纹谷线，从而获得指纹图案。
本申请实施例还提出了一种电子设备，可以包括：
处理器；
存储器，用于存储所述处理器的可执行指令；
其中，所述处理器配置为经由执行所述可执行指令来执行上述任意一项所述的指纹识别方法。
在本申请的示例性实施例中，所述电子设备还可以包括上述任意一项所述的指纹感测系统。
本申请实施例还提出了一种存储介质，所述存储介质可以包括存储的程序，其中，在所述程序运行时控制所述存储介质所在设备执行上述任意一项所述的指纹识别方法。
通过上述本申请实施例方案，减少了指纹采集次数，降低了指纹采集的繁琐度，提高了指纹图像的质量，提高了指纹解锁效率，提高了用户体验感。因此，至少包括以下有益效果：
1、只需少量次数的指纹采集便可完成身份注册，方便快捷。
2、可在手机屏幕的任意位置进行指纹识别，灵活自由。
3、在应用设计上可实现指纹识别、应用程序开启、身份验证一步到位。
4、对于扭曲变形的指纹也具有较好的识别效果，稳定可靠。
本申请实施例的其它特征和优点将在随后的说明书中阐述，并且，部分地从说明书中变得显而易见，或者通过实施本申请而了解。本申请的目的和其他优点可通过在说明书、权利要求书以及附图中所特别指出的结构来实现和获得。
附图说明
附图用来提供对本申请技术方案的进一步理解，并且构成说明书的一部分，与本申请的实施例一起用于解释本申请的技术方案，并不构成对本申请技术方案的限制。
图1为本申请实施例的指纹图像增强方法流程图；
图2为本申请实施例的采用方向场字典的方法对有效区域的初始方向场中错误的部分进行纠正的方法流程图；
图3为本申请实施例的第二预处理方法流程图；
图4为本申请实施例的指纹识别方法流程图；
图5为本申请实施例的在对所述当前帧指纹图像进行背景纹理消除以获 得纯指纹图像以后的增强处理方法流程图；
图6为本申请实施例的对经过方向场校正的有效区域进行第二预处理的具体方法流程图；
图7为本申请实施例的对指纹增强图像进行特征提取之前的方法流程图；
图8为本申请实施例的对所述扭曲指纹图像进行扭曲校正的方法流程图；图9-a为本申请实施例的脊线、细节点、子脊线的第一示意图；
图9-b为本申请实施例的脊线、细节点、子脊线的第二示意图；
图9-c为本申请实施例的子脊线的标号关系示意图；
图10为本申请实施例的基于指纹识别的应用程序启动方法流程图；
图11为本申请实施例的指纹感测系统组成结构框图；
图12为本申请实施例的指纹感测系统包括触控面板时的结构示意图；
图13为本申请实施例的通过检测自显示屏发出并在手指表面反射回显示屏的光线来获得指纹图像的方法示意图；
图14为本申请实施例的通过检测从手指透入显示屏的光线获得指纹图像的方法示意图；
图15为本申请实施例的电子设备组成结构框图。
具体实施方式
为使本申请实施例的目的、技术方案和优点更加清楚明白，下文中将结合附图对本申请的实施例进行详细说明。需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互任意组合。
在附图的流程图示出的步骤可以在诸如一组计算机可执行指令的计算机系统中执行。并且，虽然在流程图中示出了逻辑顺序，但是在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤。
为了达到本申请实施例目的，本申请实施例提供了一种指纹图像增强方法，如图1所示，所述方法可以包括S101-S105：
S101、消除当前帧指纹图像的背景纹理，获取纯指纹图像。
在本申请的示例性实施例中，在屏下指纹解锁过程中，由于屏下采集的光学指纹图像较弱，使得解锁过程中通常需要多次采集指纹图像进行验证，从而增加了用户操作的繁琐度，降低了用户体验感。为了解决这一问题，本申请实施例方案可以首先对采集到的指纹图像进行增强处理。
在本申请的示例性实施例中，所述增强处理可以包括：对所述当前帧指纹图像进行背景纹理消除以获得纯指纹图像。
在本申请的示例性实施例中，因为通常采集到的图像中不仅包含指纹，同时也包含指纹图像的背景纹理(例如，屏幕自身的纹理、残留指纹图像等)，因此在增强处理中可以先对指纹与背景纹理进行分离。
在本申请的示例性实施例中，所述消除指纹图像的背景纹理，获得纯指纹图像可以包括：
在所述当前帧指纹图像中对所述背景纹理进行像素对应消减，获取所述纯指纹图像。
在本申请的示例性实施例中，所述背景纹理可以用所述当前帧指纹图像之前N帧指纹图像的像素均值表征，其中，N为正整数。
在本申请的示例性实施例中，可以采用多帧图像求像素平均值的方法得到背景纹理，即，用当前帧指纹图像之前的N帧指纹图像序列计算像素均值，得到的结果可近似认为是背景纹理。这是由于采集到的指纹图像序列中，背景纹理相对固定且强度较强，而指纹变化较大且强度较弱，通过多帧图像平均可进一步弱化偶然出现的指纹，保留相对稳定的背景纹理。得到背景纹理后，可以采用当前帧指纹图像与背景纹理进行像素对应消减，得到去除背景纹理后的纯指纹图像。
在本申请的示例性实施例中，在所述当前帧指纹图像中对所述背景纹理进行像素对应消减可以包括：将当前帧指纹图像的像素减去该当前帧指纹图像之前N帧图像的像素平均值，得到的结果再根据特定的算法(例如乘以预设系数，再与一个预设数值相加)实现当前帧指纹图像中的背景纹理消减。
在本申请的示例性实施例中，所述的指纹图像增强方法还可以包括：在消除当前帧指纹图像的背景纹理之前，对所述当前帧指纹图像和所述背景纹理进行局部颜色传递，以使所述当前帧指纹图像和所述背景纹理的亮度保持一致。
在本申请的示例性实施例中，对所述当前帧指纹图像进行背景纹理消除可能出现的问题是当前帧指纹图像与背景纹理的亮度不一致，直接相减会产生错误的结果。因此，可以先对指纹图像与背景纹理进行局部颜色传递，使背景纹理与指纹图像的亮度保持一致，然后进行逐像素对应消减，得到相对纯净的指纹图像，即上述的纯指纹图像。
S102、对所述纯指纹图像进行第一预处理，获得第一预处理图像。
在本申请的示例性实施例中，所述第一预处理可以包括对所述纯指纹图像进行对比度增强和/或去噪。
在本申请的示例性实施例中，对所述纯指纹图像进行对比度增强和去噪可以包括：
对所述纯指纹图像进行局部对比度规范化或局部自适应直方图均衡化处理，以增强所述纯指纹图像的对比度；
采用预设的去噪算法对增强对比度后的纯指纹图像进行第一滤波。
在本申请的示例性实施例中，由于指纹图像较弱，背景纹理消减后得到的指纹图像通常对比度很低，且有时还会出现整体对比度不均匀的情况。因此，可以对背景纹理消减后的纯指纹图像采用局部对比度规范化(Local Contrast Normalization,LCN)或局部自适应直方图均衡化处理，增强图像的对比度，且使图像整体对比度相对均匀。在此过程中，图像原有的噪声会在一定程度上被放大，因此可以采用预设的去噪算法采取额外的去噪处理，对增强对比度后的指纹图像进行噪声抑制。至此，便得到了相对清晰的第一预处理图像。
在本申请的示例性实施例中，所述预设的去噪算法可以包括：快速非局部均值去噪算法(Fast Non-Local Means Denoising)。
在本申请的示例性实施例中，很多经典的去噪算法会使图像变得模糊，该快速非局部均值去噪算法能够在去噪的同时较好地保留图像中的边缘信息。另外，在本申请其他实施例中也可以采用双边滤波(Bilateral Filter)或各项异性扩散滤波(Anisotropic Filter)代替快速非局部均值去噪算法。
需要注意的是，本领域技术人员根据实际应用的需求可以采用对比度增强和去噪两个步骤，也可以仅采用对比度增强或去噪。
S103、获取所述第一预处理图像的有效区域。
在本申请的示例性实施例中，在经过前面的第一预处理之后，得到的指纹图像质量已经相对较好，然而还可以做进一步的处理。
在本申请的示例性实施例中，可以获取经过所述第一滤波之后的纯指纹图像中的有效区域；所述有效区域是根据预设的指纹前景分割算法对所述纯指纹图像计算获得的。
在本申请的示例性实施例中，通常纯指纹图像中的有效区域是指指纹图像的中间部分，指纹图像的周围区域往往是无效的背景部分，对无效部分进行处理不仅会增加时间开销，还有可能带来额外的干扰。因此，可以对指纹图像进行分割，提取有效的指纹前景，去除无效的背景部分。
在本申请的示例性实施例中，所述获取所述第一预处理图像的有效区域可以采用预设的指纹前景分割算法。
在本申请的示例性实施例中，预设的指纹前景分割算法可以包括：一种改进的基于梯度的指纹图像分割算法(Fingerprint Image Segmentation Based on Boundary Values)，或者，基于图像块的灰度均值或灰度方差的分割方法。通过预设的指纹前景分割算法可以提高后续处理的性能，同时排除不必要的干扰。在得到指纹前景区域(即上述的有效区域)后，对该区域进行规范化处理，可去除指纹采集时按压力度不同带来的图像强度差异。
S104、对所述有效区域进行方向场估计和方向场校正。
在本申请的示例性实施例中，可以对纯指纹图像中的有效区域进行方向场估计。方向场是指纹图像的一个固有属性，它定义了指纹的脊线和谷线在 局部邻域内的不变坐标。
在本申请的示例性实施例中，所述方向场估计可以采用傅里叶变换、梯度法等方法估计指纹图像的初始方向场；所述方向场校正可以包括：采用方向场字典的方法对所述有效区域的初始方向场中错误的部分进行纠正，得到更加准确的方向场结果。
在本申请的示例性实施例中，如图2所示，所述采用方向场字典的方法对所述有效区域的初始方向场中错误的部分进行纠正可以包括S201-S203：
S201、将所述有效区域划分成多个区块，并计算每个区块中的方向分布特性；
S202、将每个区块中的方向分布特性与预设的方向场字典进行比较，得到相似度参数；
S203、将相似度参数大于预设的相似性阈值的方向场字典中的区块作为参考区块，根据参考区块中的方向分布特性，校正所述有效区域中的相应区块的方向分布特性；
其中，所述预设的方向场字典是通过满足预设的质量要求的样本指纹图像进行训练，提取样本指纹图像中所有区块中的方向分布特性并进行聚类获得的。
在本申请的示例性实施例中，可以把指纹图像划分成多个区块，计算每个区块中的方向分布特性，这样每个区块就称为方向场字典中的一个单词。可以预先采用质量较好的指纹图像进行训练，提取所有区块中的方向分布特性并进行聚类，得到完整的方向场字典，字典中每个区块的方向分布都比较连续且平滑。对于当前指纹图像的初始方向场，用相同的方法划分区块并与训练获得的方向场字典进行比较，用方向场字典中最相似的区块对当前区块的方向分布进行纠正，使当前指纹图像中的方向场更加准确。
S105、对经过方向场校正的所述有效区域进行第二预处理，获得指纹增强图像。
在本申请的示例性实施例中，如图3所示，所述第二预处理可以包括 S301-S303：
S301、对所述有效区域去噪。
在本申请的示例性实施例中，可以用Gabor滤波器对经过所述方向场校正的有效区域进行所述第二滤波，以去除图像中的噪声，并保留正弦状的脊线和谷线。
S302、获取二值化图像。
在本申请的示例性实施例中，可以采用自适应图像二值化算法，对所述第二滤波后的有效区域中不同区域计算最优阈值，获得所述有效区域中指纹信息的完整的二值化图像。
S303、将所述二值化图像中的指纹脊线进行细化，获得所述指纹增强图像。
在本申请的示例性实施例中，可以将所述二值化图像中的脊线细化到一个像素的宽度，并保留指纹原有的拓扑结构且不增加额外的噪点。
在本申请的示例性实施例中，在得到较为准确的方向场信息之后，可以采用Gabor滤波器对指纹前景进行滤波，可以去除图像中的部分噪声，保留正弦状的脊线和谷线。然后对滤波后的指纹图像进行二值化操作，期间，选择合适的阈值非常重要，可以采用自适应图像二值化算法，对不同区域计算最优的阈值，获得指纹信息完整的二值化图像。指纹图像预处理的最后一步可以是细化，这一步可以把二值图中的脊线细化到一个像素的宽度，同时保留指纹原有的拓扑结构且不增加额外的噪点，以便于后续的特征提取。
为了达到本申请实施例目的，本申请实施例提供了一种指纹识别方法，如图4所示，所述方法可以包括S401-S403：
S401、对采集的当前帧指纹图像进行增强处理；所述增强处理包括：对所述当前帧指纹图像进行背景纹理消除以获得纯指纹图像。
在本申请的示例性实施例中，在屏下指纹解锁过程中，由于屏下采集的光学指纹图像较弱，使得解锁过程中通常需要多次采集指纹图像进行验证，从而增加了用户操作的繁琐度，降低了用户体验感。为了解决这一问题，本 申请实施例方案可以首先对采集到的指纹图像进行增强处理。
在本申请的示例性实施例中，所述增强处理可以包括：对所述当前帧指纹图像进行背景纹理消除以获得纯指纹图像。
在本申请的示例性实施例中，因为通常采集到的图像中不仅包含指纹，同时也包含指纹图像的背景纹理(例如，屏幕自身的纹理、残留指纹图像等)，因此在增强处理中可以先对指纹与背景纹理进行分离。
在本申请的示例性实施例中，所述对所述当前帧指纹图像进行背景纹理消除以获得纯指纹图像可以包括：
采用所述当前帧指纹图像之前的N帧指纹图像序列计算像素均值，将计算结果作为所述背景纹理；N为正整数；
根据所述背景纹理在所述当前帧指纹图像中进行像素对应消减，以获取所述纯指纹图像。
在本申请的示例性实施例中，可以采用多帧图像求像素平均值的方法得到背景纹理，即，用当前帧指纹图像之前的N帧指纹图像序列计算像素均值，得到的结果可近似认为是背景纹理。这是由于采集到的指纹图像序列中，背景纹理相对固定且强度较强，而指纹变化较大且强度较弱，通过多帧图像平均可进一步弱化偶然出现的指纹，保留相对稳定的背景纹理。得到背景纹理后，可以采用当前帧指纹图像与背景纹理进行像素对应消减，得到去除背景纹理后的纯指纹图像。
在本申请的示例性实施例中，在所述当前帧指纹图像中对所述背景纹理进行像素对应消减可以包括：将当前帧指纹图像的像素减去该当前帧指纹图像之前N帧图像的像素平均值，得到的结果再根据特定的算法(例如乘以预设系数，再与一个预设数值相加)实现当前帧指纹图像中的背景纹理消减。
在本申请的示例性实施例中，在根据所述背景纹理在所述当前帧指纹图像中进行像素对应消减之前，所述方法还可以包括：
对所述当前帧指纹图像和所述背景纹理进行局部颜色传递，以使所述当前帧指纹图像和所述背景纹理的亮度保持一致。
在本申请的示例性实施例中，对所述当前帧指纹图像进行背景纹理消除可能出现的问题是当前帧指纹图像与背景纹理的亮度不一致，直接相减会产生错误的结果。因此，可以先对指纹图像与背景纹理进行局部颜色传递，使背景纹理与指纹图像的亮度保持一致，然后进行逐像素对应消减，得到相对纯净的指纹图像，即上述的纯指纹图像。
在本申请的示例性实施例中，如图5所示，在对所述当前帧指纹图像进行背景纹理消除以获得纯指纹图像以后，所述增强处理还可以包括S501-S504：
S501、对所述纯指纹图像进行第一预处理，获得第一预处理图像。
在本申请的示例性实施例中，所述第一预处理可以包括：
对所述纯指纹图像进行局部对比度规范化或局部自适应直方图均衡化处理，以增强所述纯指纹图像的对比度；
采用预设的去噪算法对增强对比度后的纯指纹图像进行第一滤波。
在本申请的示例性实施例中，由于指纹图像较弱，背景纹理消减后得到的指纹图像通常对比度很低，且有时还会出现整体对比度不均匀的情况。因此，可以对背景纹理消减后的纯指纹图像采用局部对比度规范化(Local Contrast Normalization,LCN)或局部自适应直方图均衡化处理，增强图像的对比度，且使图像整体对比度相对均匀。在此过程中，图像原有的噪声会在一定程度上被放大，因此可以采用预设的去噪算法采取额外的去噪处理，对增强对比度后的指纹图像进行噪声抑制。至此，便得到了相对清晰的第一预处理图像。
在本申请的示例性实施例中，所述预设的去噪算法可以包括：快速非局部均值去噪算法(Fast Non-Local Means Denoising)。
在本申请的示例性实施例中，很多经典的去噪算法会使图像变得模糊，该快速非局部均值去噪算法能够在去噪的同时较好地保留图像中的边缘信息。另外，在本申请其他实施例中也可以采用双边滤波(Bilateral Filter)或各项异性扩散滤波(Anisotropic Filter)代替快速非局部均值去噪算法。
S502、获取所述第一预处理图像的有效区域。
在本申请的示例性实施例中，在经过前面的第一预处理之后，得到的指纹图像质量已经相对较好，然而还可以做进一步处理，例如，获取经过所述第一滤波之后的纯指纹图像中的有效区域；所述有效区域是根据预设的指纹前景分割算法对所述纯指纹图像计算获得的。
在本申请的示例性实施例中，通常纯指纹图像中的有效区域是指指纹图像的中间部分，指纹图像的周围区域往往是无效的背景部分，对无效部分进行处理不仅会增加时间开销，还有可能带来额外的干扰。因此，可以对指纹图像进行分割，提取有效的指纹前景，去除无效的背景部分。
在本申请的示例性实施例中，预设的指纹前景分割算法可以包括：一种改进的基于梯度的指纹图像分割算法(Fingerprint Image Segmentation Based on Boundary Values)，或者，基于图像块的灰度均值或灰度方差的分割方法。通过预设的指纹前景分割算法可以提高后续处理的性能，同时排除不必要的干扰。在得到指纹前景区域(即上述的有效区域)后，对该区域进行规范化处理，可去除指纹采集时按压力度不同带来的图像强度差异。
S503、对所述有效区域进行方向场估计和方向场校正。
在本申请的示例性实施例中，可以对纯指纹图像中的有效区域进行方向场估计。方向场是指纹图像的一个固有属性，它定义了指纹的脊线和谷线在局部邻域内的不变坐标。
在本申请的示例性实施例中，所述方向场估计可以采用傅里叶变换、梯度法等方法估计指纹图像的初始方向场；所述方向场校正可以包括：采用方向场字典的方法对所述有效区域的初始方向场中错误的部分进行纠正，得到更加准确的方向场结果。
在本申请的示例性实施例中，所述采用方向场字典的方法对所述有效区域的初始方向场中错误的部分进行纠正可以包括：
将所述有效区域划分成多个区块，并计算每个区块中的方向分布特性；
将每个区块中的方向分布特性与预设的方向场字典进行比较，得到相似度参数；
将相似度参数大于预设的相似性阈值的区块作为参考区块，根据参考区块中的方向分布特性，校正所述有效区域中的相应区块的方向分布特性；
其中，所述预设的方向场字典是通过满足预设的质量要求的样本指纹图像进行训练，提取所述样本指纹图像中所有区块中的方向分布特性并进行聚类获得的。
在本申请的示例性实施例中，可以把指纹图像划分成多个区块，计算每个区块中的方向分布特性，这样每个区块就称为方向场字典中的一个单词。可以预先采用质量较好的指纹图像进行训练，提取所有区块中的方向分布特性并进行聚类，得到完整的方向场字典，字典中每个区块的方向分布都比较连续且平滑。对于当前指纹图像的初始方向场，用相同的方法划分区块并与训练获得的方向场字典进行比较，用方向场字典中最相似的区块对当前区块的方向分布进行纠正，使当前指纹图像中的方向场更加准确。
S504、对经过方向场校正的所述有效区域进行第二预处理，获得指纹增强图像。
在本申请的示例性实施例中，所述第二预处理可以包括：第二滤波、二值化处理和细化。
在本申请的示例性实施例中，如图6所示，所述对经过所述方向场校正的有效区域进行第二预处理具体可以包括S601-S603：
S601、用Gabor滤波器对经过所述方向场校正的有效区域进行所述第二滤波，以去除图像中的噪声，并保留正弦状的脊线和谷线；
S602、采用自适应图像二值化算法，对所述第二滤波后的有效区域中不同区域计算最优阈值，获得所述有效区域中指纹信息的完整的二值化图像；
S603、将所述二值化图像中的脊线细化到一个像素的宽度，并保留指纹原有的拓扑结构且不增加额外的噪点。
在本申请的示例性实施例中，在得到较为准确的方向场信息之后，可以采用Gabor滤波器对指纹前景进行滤波，可以去除图像中的部分噪声，保留正弦状的脊线和谷线。然后对滤波后的指纹图像进行二值化操作，期间，选 择合适的阈值非常重要，可以采用自适应图像二值化算法，对不同区域计算最优的阈值，获得指纹信息完整的二值化图像。指纹图像预处理的最后一步可以是细化，这一步可以把二值图中的脊线细化到一个像素的宽度，同时保留指纹原有的拓扑结构且不增加额外的噪点，以便于后续的特征提取。
在本申请的示例性实施例中，如图7所示，在对所述指纹增强图像进行特征提取之前还可以包括S701-S702：
S701、对所述指纹增强图像进行指纹扭曲检测，确定所述指纹增强图像为正常指纹图像或扭曲指纹图像；
S702、对所述扭曲指纹图像进行扭曲校正。
在本申请的示例性实施例中，在指纹识别过程中通常还会遇到指纹扭曲变形的问题，这是由于采集过程中手指按压的力度和方向不同造成的，这会导致同一手指产生不同的特征数据，影响最后的识别结果。因此，可以采用指纹扭曲检测和校正算法，把发生扭曲的指纹图像校正到未扭曲变形的状态，从而保证最后得到的特征数据的一致性。
在本申请的示例性实施例中，所述对所述指纹增强图像进行指纹扭曲检测可以包括：
将所述指纹增强图像输入分类器进行分类，分类结果包括正常指纹图像和扭曲指纹图像。
在本申请的示例性实施例中，指纹的扭曲变形会导致最终提取到的特征数据与正常状态不同，极大降低匹配的分数，造成错误的识别结果。因此可以先对指纹图像进行扭曲检测，如果检测到扭曲变形，就对该指纹图像进行校正，使指纹图像恢复到正常的状态。为此，可以采用事先采集的大量正常指纹图像及扭曲指纹图像训练出一个分类器，将经过增强处理的指纹图像输入该训练好的分类器中，对当前输入的指纹图像做一个二分的分类，如果分类结果属于扭曲，则对该指纹图像进行扭曲校正。
在本申请的示例性实施例中，如图8所示，所述对所述扭曲指纹图像进行扭曲校正可以包括S801-S803：
S801、提取所述扭曲指纹图像的方向场和周期图；
S802、根据所述方向场和周期图在数据库集中查找与所述扭曲指纹图像最接近的参考扭曲指纹；
S803、根据所述参考扭曲指纹对所述扭曲指纹图像进行逆变换校正。
在本申请的示例性实施例中，扭曲校正可以通过估计扭曲指纹图像的扭曲场并对其逆变换来完成。为此，我们构建了一个数据库集，里面包含了各种扭曲指纹对应的扭曲场(扭曲场是指一个指纹从正常非扭曲状态到扭曲状态之间的变换关系)、方向场及周期图(周期图是指指纹图像中各个位置的脊线周期或频率(代表脊线的密集程度))。具体方法可以包括：采集常见的正常指纹与扭曲指纹的图像对，从而通过这些图像对获得常见扭曲场的统计模型，采用这些统计模型合成大量扭曲场并作用到正常指纹图像上，进而获得正常指纹图像对应的扭曲指纹图像及它们的方向场和周期图的集合，作为上述的数据库集。
在本申请的示例性实施例中，对指纹图像的扭曲校正可以包括：对于当前检测到的扭曲指纹图像，可以先提取它的方向场和周期图，然后在数据库集中查找与当前扭曲指纹图像特征最接近的参考扭曲指纹，根据该扭曲指纹对应的扭曲场对当前扭曲指纹图像进行逆变换校正，使当前扭曲指纹图像恢复正常状态。
S402、对所述指纹增强图像进行特征提取，获取特征数据。
在本申请的示例性实施例中，所述特征数据可以包括但不限于：指纹脊线的细节点特征、脊线特征。
在本申请的示例性实施例中，通过上述步骤以后便可以完成指纹图像增强处理，并可以完成扭曲指纹的校正，从而获得高质量的指纹图像。针对该高质量的指纹图像，便可对细化的指纹脊线提取特征，以获取特征数据。
在本申请的示例性实施例中，可以提取细节点特征，该细节点特征可以包括脊线的端点和分叉点等。通过对细节点特征进行预设形式的编码存储，就可以得到了当前指纹图像的特征数据。
S403、根据所述特征数据与指纹模板的特征数据的比对完成指纹识别。在本申请的示例性实施例中，所述根据所述特征数据与指纹模板的特征数据的比对完成指纹识别可以包括：
计算所述特征数据与所述指纹模板的特征数据的特征相似度；
当所述特征相似度大于或等于特征相似度阈值时，完成指纹识别。
在本申请的示例性实施例中，将获得的当前指纹图像的特征数据与来自指纹模板的不同指纹的特征数据进行比对，就可以计算出不同指纹之间的相似度，从而根据该相似度可以完成指纹识别与验证。通常情况下，当前指纹图像的所述特征数据与预设的不同指纹的特征数据中的任何一个的相似度大于或等于预设的相似度阈值时，确定两者匹配，并可以确定验证成功。
在本申请的示例性实施例中，通常情况下，进行指纹识别最后阶段，进行指纹比对时，内部存储的指纹模板(即预设的不同指纹的特征数据)与当前输入指纹的姿态不同，无法直接进行匹配，因此可以首先对两幅相互比对的指纹图像进行对齐。指纹对齐通常是通过寻找最相似的子结构(比如细节点集或脊线)完成的，在此，采用的子结构可以是细节点及多条相关的脊线，具体为：细节点、细节点所在的脊线及该脊线两侧相邻的脊线所构成的子结构。对于该子结构，先确定细节点(脊线内部的端点或分叉点表示细节点，如图9-a、9-b中黑点所示)，再通过细节点沿垂直脊线方向画一条直线，直线与相邻脊线的两个交点称为投影点，通过细节点及投影点将子结构中的脊线进行拆分，并根据它们的相对位置及方向进行标号，如图9-a、图9-b所示。图中原始的完整黑线表示脊线(如图9-a中的2+3、4+5、1和图9-b中4+5、3+1、3+2、6+7)，脊线内部的端点或分叉点表示细节点(如图9-a、9-b中黑点所示)，图中白点为投影点，脊线被细节点或投影点拆分得到子脊线(如9-a中的2、3、1、4、5和图9-b中的4、5、3、1、2、6、7所示)。
在本申请的示例性实施例中，如果两个子结构满足以下条件中的任意一个或多个，便可以判定它们相互匹配：1.细节点类型相同，且对应的子脊线标号相同；2.细节点类型不同，但子脊线的标号关系为给定的两种之一(如图9-c所示)；3.对应的子脊线的相似度大于一定的相似度阈值，且所有子 脊线的平均相似度大于一定的相似度阈值。接着，可以从所有匹配的子结构中选出最匹配的前N对，通过最小二乘法估计出它们之间整体的仿射变换，从而对两幅指纹图像完成对齐。
在本申请的示例性实施例中，理想情况下，原本匹配的细节点对及子结构对经过对齐变换后应该是完全重叠的，但实际上由于细节点提取过程的误差以及对齐变换与真实物理变换的误差，导致细节点对与子结构对无法完全匹配。因此可以采用一种更加稳定的匹配方案来计算指纹间的相似度，可以主要考虑到以下两个方面：1.细节点对方面：可以首先选取一个参考细节点，将其他所有细节点转换成以参考细节点为原点的极坐标；然后可以按照角度升序的方式将所有细节点连接组成一个特征字符串；最后可以计算指纹模板特征字符串与当前指纹特征字符串之间的编辑距离，根据编辑距离确定细节点对之间的匹配分数。2.子结构对方面：可以对指纹对齐时得到的最匹配的前N对子结构进行遍历，每一对子结构中的对应脊线构成初始的匹配脊线对，匹配脊线对相邻的脊线对又构成新的匹配脊线对，由此得到两个指纹的匹配脊线对集合。对任意一对匹配的脊线对，可以通过它们所在子结构对的变换或生成它们的脊线对所在子结构对的变换进行对齐，然后可以通过动态规划计算出最佳匹配序列，得到匹配脊线对之间的匹配点数。同时，对于子结构对中的细节点，如果它们距离所在匹配脊线对中的匹配点小于一定的距离阈值，且细节点相邻脊线之间的周期与对应细节点相邻脊线之间的周期比较接近(如小于预设的差异阈值)，则可以认为细节点对相互匹配。最后，可以通过匹配脊线对上的匹配点的比例及匹配细节点占匹配子结构对的比例，得到子结构对之间的匹配分数。可以把细节点对匹配分数与子结构对匹配分数综合起来，得到两幅指纹图像之间最终的相似度，通过将该做种的相似度与预设的相似度阈值比较可以确认当前的两个指纹是否匹配成功。
根据本申请实施例的另一方面，还提供了一种基于指纹识别的应用程序启动方法，如图10所示，该应用程序启动方法还可以包括S1001-S1003：
S1001、采集指纹图像；
S1002、根据上述任意一项指纹识别方法，对指纹图像进行识别；
S1003、当指纹识别成功时开启该应用程序。
在本申请的示例性实施例中，应用程序启动方法还可以包括在所述采集指纹图像之前，检测触控屏上手指的触控操作是否满足启动采集指纹图像步骤的预设条件。
在本申请的示例性实施例中，应用程序可以为仅允许授权人员访问以保护用户隐私、个人信息或企事业单位机密信息的计算机程序。
在本申请的示例性实施例中，应用程序启动方法还可以在开启应用程序的同时完成身份验证。
在本申请的示例性实施例中，可以在指纹识别成功开启应用程序的同时收集该应用程序和该指纹对应的用户身份信息，用于大数据分析。例如，可以通过收集用户使用应用程序的频率、时间点等信息分析用户的偏好和习惯，由此，可以有助于应用程序开发者进行市场规划。
在传统的应用程序启动方法中，首先要通过手指点击或触摸开启应用程序，然后在应用程序提示要指纹验证时，再次通过手指按压在特定的指纹识别区域，才能完成身份验证。可见，在这种传统的方法中，至少需要两个步骤才能开启应用程序并完成身份验证，操作较为繁琐且耗时，在一定程度上降低了用户的体验度。而在本申请实施例中，通过上述步骤，在手指点击、触摸或接近屏幕时，即可在启动应用程序的同时完成身份验证，以确认用户有权限对应用程序进行相应的操作，提供对应用程序的安全访问(例如，安全金融交易)。
根据本申请实施例的另一方面，还提供了一种指纹感测系统1，如图11所示，所述指纹感测系统可以包括：显示屏11、指纹采集模组12和指纹识别模13。
显示屏11，包括发光显示单元，用于显示画面。
在本申请的示例性实施例中，发光显示单元可以为自发光显示单元，例如发光二极管(Light-Emitting Diode,LED)、有机发光二极管(Organic Light-Emitting Diode,OLED)或微型发光二极管(Micro-LED)等。在其他替 代实施例中，发光显示单元也可以为被动发光显示单元，例如液晶显示屏(Liquid Crystal Display,LCD)等。
指纹采集模组12，至少设置在所述显示屏下方的局部区域，用于采集指纹图像。
在本申请的示例性实施例中，由于便携式或可穿戴式设备的空间有限，通常希望尽可能将显示区域最大化，因此，可以将指纹采集模组至少设置在显示屏下方的局部区域，以减少对显示区域的占用。
指纹识别模13，用于接收所述当前帧指纹图像，并采用上述指纹识别方法，对所述指纹图像进行指纹识别。
在本申请的示例性实施例中，显示屏11可以为触控显示屏，不仅可以进行画面显示，还可以检测用户手指的操作(例如触摸、按压或接近显示屏)，从而为用户提供一个人机交互的界面。例如，在一个实施例中，如图12所示，指纹感测系统还可以包括触控面板(Touch Panel,TP)，该触控面板可以设置在显示屏的表面，也可以部分或整体集成于显示屏中，构成触控显示屏。
在本申请的示例性实施例中，指纹感测系统1还可以包括盖板，设置在显示屏的上方，作为用户触摸和显示画面的界面，以实现对显示屏的保护。盖板可以为玻璃或蓝宝石，而且不限于此。
在本申请的关于指纹采集模组12的组成结构的第一个示例性实施例中，指纹采集模组12可以包括光学准直器和光电检测器(Photo Detector)。通过光学准直器，只有入射角小于预设角的光线才能到达光电检测器。
在本申请的关于指纹采集模组12的组成结构的第二个示例性实施例中，指纹采集模组可以包括透镜和成像单元，其中，成像单元可以设置于透镜下方，利用透镜成像原理直接获取显示屏上的指纹图像。其中，透镜可以选用凸透镜。根据实际应用的需求，可以在显示器下方分别设置一个或多个透镜和成像单元，从而实现在屏幕的局部区域、半屏或全屏的指纹采集及识别。透镜和成像单元可以为相互独立的部件或集成为一体的部件，当透镜和成像单元相互独立时，两者的数量不一定为一一对应。
在本申请的关于指纹采集模组12的组成结构的第三个示例性实施例中，所述指纹采集模组用于通过检测自显示屏发出并在手指表面反射回显示屏的光线来获得指纹图像。具体地，指纹采集模组可以包括第一光电传感器，第一光电传感器可以通过检测自显示屏发出并在手指表面反射回显示屏的光线来获得指纹图像。如图13所示，其中，在距离显示屏中某个点亮的发光显示单元足够远的地方，指纹脊线处的折射率大于指纹谷线处的折射率，因此在指纹谷线处会形成全反射，在指纹脊线处却不会形成全反射，而是有一部分光会透射到手指内部，由此会导致在指纹谷线处反射光的强度大于指纹脊线处的反射光，通过检测这些反射回显示屏的光线的强弱，可以确定指纹脊线及指纹谷线，从而获得指纹图像。但是，在这种方案中，由于激励光为内置的发光显示单元，且要检测反射光的强弱，环境光对效果的影响较大，当反射光强弱差异较小时，会降低指纹图像的清晰度。
在本申请的关于指纹采集模组12的组成结构的第四个示例性实施例中，如图14所示，所述指纹采集模组12还可以通过检测从手指透入显示屏的光线获得指纹图像，其中，当检测到折射角大于第一阈值的光线时，可以将该光线确定为指纹脊线，否则将该光线确定为指纹谷线，根据指纹脊线和指纹谷线获得所述指纹图像。所述第一阈值可以为所述指纹谷线处的折射角。优选的，指纹采集模组12可以包括光路引导模组，用于引导折射角大于第一阈值的光线。所述指纹采集模组12还可以包括光电传感器(第二光电传感器)，第二光电传感器可以用于在检测到折射角大于第一阈值的光线时，确定为指纹脊线，否则为指纹谷线，从而获得指纹图案。第二光电传感器可以为互补金属氧化物半导体CMOS传感器、薄膜晶体管TFT传感器或其他定制的传感器。在本实施例中，由于激励光本身就是环境光，故指纹图像的获得不受环境光的影响，理论上，环境光越强，采集效果越好。此外，当环境光不足时，可以通过点亮手指周围的显示屏内置的发光显示单元或通过点亮外部光源，以增强光线的强度。
通过上述本申请的示例性实施例可知，由于指纹感测系统是位于显示屏下方的，显示屏的显示区域可以扩展到电子设备的整个面。并且，指纹感测系统可以设置在显示屏下方的部分区域或全部区域，从而实现局部区域、半 屏或全屏的指纹识别。在现有技术中，指纹感测系统通常设置在显示屏外的区域且与手指的接触面较小，例如苹果iphone 6，对识别对象和识别方式有极大的限制，不适合采集识别大图案(例如，掌纹)，也不适合对多个识别对象同时进行识别验证。而采用本申请实施例的技术方案，由于在显示屏下方的局部区域、半屏区域或全屏区域均可以设置一个或多个独立或整体的指纹感测系统，因此可以实现对大图案(例如，掌纹)的采集识别，以扩大应用场景；并且可以实现对多个识别对象同时进行识别验证，以增强应用程序的安全性，例如可以设置某个金融支付类应用程序需要同时验证两个人的指纹才能开启并完成支付，则可以通过设置全屏的指纹识别，使得两个人的手指可以同时触摸显示屏的任何区域。
在本申请的示例性实施例中，本申请实施例的技术方案可以应用于各种具有显示屏的电子设备，例如智能手机、笔记本电脑、平板电脑、数码相机、游戏机、智能手环、智能手表等便携式或穿戴式电子设备，以及自动柜员机(Automated Teller Machine，ATM)、信息管理系统、电子门锁等其它电子设备。
在本申请的示例性实施例中，本申请实施例的技术方案还可以进行除指纹以外的其它生物特征识别，本申请实施例对此不作限定。
根据本申请实施例的另一方面，还提供了一种电子设备A，如图15所示，该电子设备可以包括：
处理器2；以及
存储器3，用于存储所述处理器2的可执行指令；
其中，所述处理器2配置为经由执行所述可执行指令来执行前述任意一项所述的指纹识别方法。
在本申请的示例性实施例中，如图15所示，电子设备A还可以包括上述任意一项所述的指纹感测系统1。
根据本申请实施例的另一方面，还提供了一种存储介质，所述存储介质包括存储的程序，其中，在所述程序运行时控制所述存储介质所在设备执行 前述内容中任意一项所述的指纹识别方法。
本申请实施例至少包括以下有益效果：
1、只需少量次数的指纹采集便可完成身份注册，方便快捷。
2、可在手机屏幕的任意位置进行指纹识别，灵活自由。
3、在应用设计上可实现指纹识别、应用程序开启、身份验证一步到位。
4、对于扭曲变形的指纹也具有较好的识别效果，稳定可靠。
本领域普通技术人员可以理解，上文中所公开方法中的全部或某些步骤、系统、装置中的功能模块/单元可以被实施为软件、固件、硬件及其适当的组合。在硬件实施方式中，在以上描述中提及的功能模块/单元之间的划分不一定对应于物理组件的划分；例如，一个物理组件可以具有多个功能，或者一个功能或步骤可以由若干物理组件合作执行。某些组件或所有组件可以被实施为由处理器，如数字信号处理器或微处理器执行的软件，或者被实施为硬件，或者被实施为集成电路，如专用集成电路。这样的软件可以分布在计算机可读介质上，计算机可读介质可以包括计算机存储介质(或非暂时性介质)和通信介质(或暂时性介质)。如本领域普通技术人员公知的，术语计算机存储介质包括在用于存储信息(诸如计算机可读指令、数据结构、程序模块或其他数据)的任何方法或技术中实施的易失性和非易失性、可移除和不可移除介质。计算机存储介质包括但不限于RAM、ROM、EEPROM、闪存或其他存储器技术、CD-ROM、数字多功能盘(DVD)或其他光盘存储、磁盒、磁带、磁盘存储或其他磁存储装置、或者可以用于存储期望的信息并且可以被计算机访问的任何其他的介质。此外，本领域普通技术人员公知的是，通信介质通常包含计算机可读指令、数据结构、程序模块或者诸如载波或其他传输机制之类的调制数据信号中的其他数据，并且可包括任何信息递送介质。

Claims (30)

  1. 一种指纹图像增强方法，所述方法包括：
    消除当前帧指纹图像的背景纹理，获取纯指纹图像；
    对所述纯指纹图像进行第一预处理，获得第一预处理图像；
    获取所述第一预处理图像的有效区域；
    对所述有效区域进行方向场估计和方向场校正；
    对经过方向场校正的所述有效区域进行第二预处理，获得指纹增强图像。
  2. 根据权利要求1所述的指纹图像增强方法，其中，所述背景纹理为所述当前帧指纹图像之前N帧指纹图像的像素均值，其中，N为正整数。
  3. 根据权利要求1所述的指纹图像增强方法，其中，所述消除指纹图像的背景纹理，获得纯指纹图像包括：
    在所述当前帧指纹图像中对所述背景纹理进行像素对应消减，获取所述纯指纹图像。
  4. 根据权利要求1所述的指纹图像增强方法，其中，在消除当前帧指纹图像的背景纹理之前，对所述当前帧指纹图像和所述背景纹理进行局部颜色传递。
  5. 根据权利要求1所述的指纹图像增强方法，其中，所述第一预处理包括对所述纯指纹图像进行对比度增强和/或去噪。
  6. 根据权利要求1所述的指纹图像增强方法，其中，所述获取所述第一预处理图像的有效区域采用预设的指纹前景分割算法。
  7. 根据权利要求1所述的指纹图像增强方法，其中，所述第二预处理包括：
    对所述有效区域去噪；
    获取二值化图像；
    将所述二值化图像中的指纹脊线进行细化，获得所述指纹增强图像。
  8. 一种指纹识别方法，所述方法包括：
    对采集的当前帧指纹图像进行增强处理，获得指纹增强图像；所述增强处理包括：消除当前帧指纹图像的背景纹理，获得纯指纹图像；
    对所述指纹增强图像进行特征提取，获取特征数据；
    根据所述特征数据与指纹模板的特征数据的比对完成指纹识别。
  9. 根据权利要求8所述的指纹识别方法，其中，所述增强处理还包括：
    对所述纯指纹图像进行第一预处理，获得第一预处理图像；
    获取所述第一预处理图像的有效区域；
    对所述有效区域进行方向场估计和方向场校正；
    对经过方向场校正的所述有效区域进行第二预处理，获得指纹增强图像。
  10. 根据权利要求8所述的指纹识别方法，其中，
    所述背景纹理为所述当前帧指纹图像之前N帧指纹图像的像素均值，其中，N为正整数。
  11. 根据权利要求8所述的指纹识别方法，其中，所述消除指纹图像的背景纹理，获得纯指纹图像包括：
    在所述当前帧指纹图像中对所述背景纹理进行像素对应消减，获取所述纯指纹图像。
  12. 根据权利要求8所述的指纹识别方法，其中，在消除当前帧指纹图像的背景纹理之前，对所述当前帧指纹图像和所述背景纹理进行局部颜色传递。
  13. 根据权利要求8所述的指纹识别方法，其中，在对所述指纹增强图像进行特征提取之前还包括：
    对所述指纹增强图像进行指纹扭曲检测，确定所述指纹增强图像为正常指纹图像或扭曲指纹图像；
    对所述扭曲指纹图像进行扭曲校正。
  14. 根据权利要求13所述的指纹识别方法，其中，对所述指纹增强图像进行指纹扭曲检测包括：
    将所述指纹增强图像输入分类器进行分类，分类结果包括正常指纹图像和扭曲指纹图像。
  15. 根据权利要求13所述的指纹识别方法，其中，对所述扭曲指纹图像进行扭曲校正包括：
    提取所述扭曲指纹图像的方向场和周期图；
    根据所述方向场和周期图在数据库集中查找与所述扭曲指纹图像最接近的参考扭曲指纹；
    根据所述参考扭曲指纹对所述扭曲指纹图像进行逆变换校正。
  16. 根据权利要求8所述的指纹识别方法，其中，所述特征数据包括指纹脊线的细节点特征和/或脊线特征。
  17. 根据权利要求8所述的指纹识别方法，其中，根据所述特征数据与指纹模板的特征数据的比对完成指纹识别包括：
    计算所述特征数据与所述指纹模板的特征数据的特征相似度；
    当所述特征相似度大于或等于特征相似度阈值时，完成指纹识别。
  18. 一种基于指纹识别的应用程序启动方法，所述应用程序启动方法包括：
    采集指纹图像；
    采用权利要求8至17中任意一项所述的指纹识别方法，对所述指纹图像进行指纹识别；
    当指纹识别成功时开启所述应用程序。
  19. 根据权利要求18所述的应用程序启动方法，其中，在所述采集指纹图像之前，检测触控屏上手指的触控操作是否满足启动采集指纹图像步骤的预设条件。
  20. 根据权利要求18所述的应用程序启动方法，其中，所述方法还包括：在开启所述应用程序的同时完成身份验证。
  21. 一种指纹感测系统，包括：
    显示屏，包括发光显示单元，设置为显示画面；
    指纹采集模组，至少设置在所述显示屏下方的局部区域，设置为采集指纹图像；
    指纹识别模组，设置为接收所述指纹图像，并采用权利要求8至17中任意一项所述的指纹识别方法，对所述指纹图像进行指纹识别。
  22. 根据权利要求21所述的指纹感测系统，其中，所述指纹采集模组包括：
    透镜；
    成像单元，设置于所述透镜下方，设置为直接获取所述显示屏上的指纹图像。
  23. 根据权利要求21所述的指纹感测系统，其中，所述指纹采集模组用于通过检测自显示屏发出并在手指表面反射回显示屏的光线来获得指纹图像。
  24. 根据权利要求21所述的指纹感测系统，其中，所述指纹采集模组通过检测从手指透入显示屏的光线获得指纹图像；其中，当检测到折射角大于第一阈值的光线时，确定为指纹脊线，当检测到折射角小于或等于所述第一阈值的光线时，确定为指纹谷线，根据所述指纹脊线和所述指纹谷线获得所述指纹图像。
  25. 根据权利要求24所述的指纹感测系统，其中，所述第一阈值为所述指纹谷线处的折射角。
  26. 根据权利要求24所述的指纹感测系统，其中，所述指纹采集模组还包括光路引导模组，设置为引导折射角大于所述第一阈值的光线。
  27. 根据权利要求24所述的指纹感测系统，其中，所述指纹采集模组还包括光电传感器，设置为在检测到折射角大于所述第一阈值的光线时，确定所述光线为指纹脊线，在检测到折射角小于或等于所述第一阈值的光线时，确定所述光线为指纹谷线，从而获得指纹图案。
  28. 一种电子设备，包括：
    处理器；以及
    存储器，用于存储所述处理器的可执行指令；
    其中，所述处理器配置为经由执行所述可执行指令来执行权利要求8至17中任意一项所述的指纹识别方法。
  29. 根据权利要求28所述的电子设备，其中，还包括权利要求20-27任意一项所述的指纹感测系统。
  30. 一种存储介质，所述存储介质包括存储的程序，其中，在所述程序运行时控制所述存储介质所在设备执行权利要求8至17中任意一项所述的指纹识别方法。
PCT/CN2020/078648 2019-03-15 2020-03-10 一种指纹图像增强、指纹识别和应用程序启动方法 WO2020187098A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
EP20774228.9A EP3940583A4 (en) 2019-03-15 2020-03-10 PROCEDURE FOR FINGERPRINT IMAGE ENHANCEMENT, FINGERPRINT RECOGNITION AND LAUNCHING AN APPLICATION
KR1020217033357A KR20210136127A (ko) 2019-03-15 2020-03-10 지문 이미지 강조, 지문 인식, 애플리케이션 프로그램 기동 방법, 지문 감지 시스템, 전자기기 및 저장 매체
US17/041,465 US11874907B2 (en) 2019-03-15 2020-03-10 Method for enhancing fingerprint image, identifying fingerprint and starting-up application program
JP2021555011A JP7377879B2 (ja) 2019-03-15 2020-03-10 指紋画像強調、指紋認識、アプリケーションプログラム起動方法、指紋検知システム、電子機器及び記憶媒体

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910198302.7A CN111695386B (zh) 2019-03-15 2019-03-15 一种指纹图像增强、指纹识别和应用程序启动方法
CN201910198302.7 2019-03-15

Publications (1)

Publication Number Publication Date
WO2020187098A1 true WO2020187098A1 (zh) 2020-09-24

Family

ID=72475347

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/078648 WO2020187098A1 (zh) 2019-03-15 2020-03-10 一种指纹图像增强、指纹识别和应用程序启动方法

Country Status (6)

Country Link
US (1) US11874907B2 (zh)
EP (1) EP3940583A4 (zh)
JP (1) JP7377879B2 (zh)
KR (1) KR20210136127A (zh)
CN (1) CN111695386B (zh)
WO (1) WO2020187098A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112884756A (zh) * 2021-03-12 2021-06-01 敦泰电子（深圳）有限公司 一种指纹图像处理方法、指纹芯片及电子设备
CN113570699A (zh) * 2021-06-24 2021-10-29 清华大学 三维指纹重建的方法和装置

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112668399B (zh) * 2020-12-07 2022-05-17 北京极豪科技有限公司 图像处理方法、指纹信息提取方法、装置、设备及介质
CN114648548A (zh) * 2020-12-17 2022-06-21 虹软科技股份有限公司 图像处理方法、描述子提取方法及其装置、电子设备
TWI785443B (zh) * 2020-12-25 2022-12-01 大陸商北京集創北方科技股份有限公司 大面積屏下光學指紋採集方法、指紋辨識裝置、及資訊處理裝置
CN113408416B (zh) * 2021-06-18 2023-04-07 展讯通信（上海）有限公司 指纹频率估计方法及装置、指纹信息提取方法及装置
CN114356103A (zh) * 2022-01-30 2022-04-15 清华大学 基于指纹图像的三维位姿增量控制方法及装置
CN114368692B (zh) * 2022-03-21 2022-06-14 杭州杰牌传动科技有限公司 智能塔吊物联网协同方法和系统
CN114913555A (zh) * 2022-05-30 2022-08-16 深圳芯启航科技有限公司 指纹特征点获取方法、装置、电子设备及存储介质
US11972632B1 (en) * 2023-03-31 2024-04-30 Republic Of Korea (National Service Director Ministry Of The Interior And Safety) Apparatus and method of digital enhancement of fingerprint ridges

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1480898A (zh) * 2003-07-21 2004-03-10 上海海运学院 指纹识别系统
CN101609499A (zh) * 2008-06-20 2009-12-23 南京理工大学 快速指纹识别方法
CN102999750A (zh) * 2012-12-31 2013-03-27 清华大学 一种去除背景干扰的现场指纹增强方法
CN103065134A (zh) * 2013-01-22 2013-04-24 江苏超创信息软件发展股份有限公司 一种具有提示信息的指纹识别装置和方法
CN106342325B (zh) * 2009-12-23 2013-08-14 凯迈(洛阳)电子有限公司 一种指纹图像的区域分割方法
CN103996026A (zh) * 2014-05-15 2014-08-20 清华大学 指纹特征提取方法、装置及系统
CN105205442A (zh) * 2015-08-07 2015-12-30 北京天诚盛业科技有限公司 指纹采集的方法和装置

Family Cites Families (26)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5963656A (en) * 1996-09-30 1999-10-05 International Business Machines Corporation System and method for determining the quality of fingerprint images
US5995642A (en) * 1997-06-30 1999-11-30 Aetex Biometric Corporation Method for automatic fingerprint classification
US7054470B2 (en) * 1999-12-02 2006-05-30 International Business Machines Corporation System and method for distortion characterization in fingerprint and palm-print image sequences and using this distortion as a behavioral biometrics
US6876757B2 (en) 2001-05-25 2005-04-05 Geometric Informatics, Inc. Fingerprint recognition system
US20040175023A1 (en) * 2001-07-05 2004-09-09 Ola Svedin Method and apparatus for checking a person's identity, where a system of coordinates, constant to the fingerprint, is the reference
JP2003051012A (ja) * 2001-08-03 2003-02-21 Nec Corp ユーザ認証方法及びユーザ認証装置
US7680354B2 (en) * 2006-03-22 2010-03-16 Arcsoft, Inc. Image deblur based on two images
CN100412883C (zh) * 2006-03-23 2008-08-20 北京中控科技发展有限公司 指纹识别方法与系统
US20080016371A1 (en) * 2006-07-14 2008-01-17 Arachnoid Biometrics Identification Group Corp. System and Method for Registering a Fingerprint, for Setting a Login Method of an Application, and for Logining in the Application
JP2008276758A (ja) 2007-03-31 2008-11-13 Univ Waseda 指紋特異点抽出装置及び指紋特異点抽出方法
GB2450479A (en) * 2007-06-22 2008-12-31 Warwick Warp Ltd Fingerprint recognition including preprocessing an image by justification and segmentation before plotting ridge characteristics in feature space
CN101079102A (zh) * 2007-06-28 2007-11-28 中南大学 基于统计方法的指纹识别方法
US20120127179A1 (en) * 2010-11-19 2012-05-24 Nokia Corporation Method, apparatus and computer program product for user interface
CN102270297B (zh) * 2011-07-21 2012-12-19 中国人民解放军国防科学技术大学 一种指纹图像增强方法
CN102609705A (zh) * 2012-01-20 2012-07-25 西安电子科技大学 一种基于扩散方程的低质量指纹图像方向场提取方法
CN102682428B (zh) * 2012-04-18 2014-11-05 浙江大学城市学院 一种基于方向场的指纹图像计算机自动化修补方法
KR101432011B1 (ko) * 2013-03-22 2014-08-21 주식회사 슈프리마 지문 영상 처리 장치 및 방법
CN104156707B (zh) * 2014-08-14 2017-09-22 深圳市汇顶科技股份有限公司 指纹识别方法及其指纹识别装置
CN104765994A (zh) * 2015-04-17 2015-07-08 努比亚技术有限公司 用户身份识别方法及装置
US10067229B2 (en) * 2015-09-24 2018-09-04 Qualcomm Incorporated Receive-side beam forming for an ultrasonic image sensor
JP2017138858A (ja) 2016-02-04 2017-08-10 富士通株式会社 生体認証装置、生体認証方法および生体認証プログラム
CN106709450A (zh) * 2016-12-23 2017-05-24 上海斐讯数据通信技术有限公司 一种指纹图像识别方法及系统
CN108664177B (zh) 2017-03-29 2021-11-12 上海耕岩智能科技有限公司 一种基于指纹识别开启应用的方法和装置
JP6948146B2 (ja) 2017-04-18 2021-10-13 サンデン・オートモーティブクライメイトシステム株式会社 車両用空気調和装置
CN107239737A (zh) * 2017-05-03 2017-10-10 广东欧珀移动通信有限公司 一种光学指纹识别方法及相关产品
CN107454963B (zh) * 2017-06-16 2021-11-30 深圳市汇顶科技股份有限公司 指纹图像处理方法及光学指纹辨识系统

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1480898A (zh) * 2003-07-21 2004-03-10 上海海运学院 指纹识别系统
CN101609499A (zh) * 2008-06-20 2009-12-23 南京理工大学 快速指纹识别方法
CN106342325B (zh) * 2009-12-23 2013-08-14 凯迈(洛阳)电子有限公司 一种指纹图像的区域分割方法
CN102999750A (zh) * 2012-12-31 2013-03-27 清华大学 一种去除背景干扰的现场指纹增强方法
CN103065134A (zh) * 2013-01-22 2013-04-24 江苏超创信息软件发展股份有限公司 一种具有提示信息的指纹识别装置和方法
CN103996026A (zh) * 2014-05-15 2014-08-20 清华大学 指纹特征提取方法、装置及系统
CN105205442A (zh) * 2015-08-07 2015-12-30 北京天诚盛业科技有限公司 指纹采集的方法和装置

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112884756A (zh) * 2021-03-12 2021-06-01 敦泰电子（深圳）有限公司 一种指纹图像处理方法、指纹芯片及电子设备
CN113570699A (zh) * 2021-06-24 2021-10-29 清华大学 三维指纹重建的方法和装置

Also Published As

Publication number Publication date
US20220084319A1 (en) 2022-03-17
JP2022524833A (ja) 2022-05-10
CN111695386A (zh) 2020-09-22
CN111695386B (zh) 2024-04-26
US11874907B2 (en) 2024-01-16
EP3940583A1 (en) 2022-01-19
EP3940583A4 (en) 2022-09-07
KR20210136127A (ko) 2021-11-16
JP7377879B2 (ja) 2023-11-10

Similar Documents

Publication Publication Date Title
WO2020187098A1 (zh) 一种指纹图像增强、指纹识别和应用程序启动方法
US10339362B2 (en) Systems and methods for performing fingerprint based user authentication using imagery captured using mobile devices
US20220165087A1 (en) Systems and methods for performing fingerprint based user authentication using imagery captured using mobile devices
US11263432B2 (en) Systems and methods for performing fingerprint based user authentication using imagery captured using mobile devices
AU2017370720B2 (en) Systems and methods for performing fingerprint based user authentication using imagery captured using mobile devices
US10558841B2 (en) Method and apparatus for recognizing fingerprint ridge point
US9710691B1 (en) Touchless fingerprint matching systems and methods
US20170154202A1 (en) Feature extraction and matching for biometric authentication
US20160379042A1 (en) Apparatuses, systems, and methods for confirming identity
CN113614731A (zh) 使用软生物特征的认证验证
Ravi et al. A novel method for touch-less finger print authentication
WO2018107574A1 (zh) 一种对印防伪特征的检测方法及装置
CN111259757A (zh) 一种基于图像的活体识别方法、装置及设备
WO2023028947A1 (zh) 掌静脉非接触式三维建模方法、装置及认证方法
US11120285B2 (en) Intelligent terminal
US11120284B2 (en) Startup authentication method for intelligent terminal
CN112801034A (zh) 一种指静脉识别装置
CN115841684A (zh) 指纹滑动录入方法、电子设备和计算机可读介质
CN112861756A (zh) 一种指静脉识别方法及系统
Galdi Design and development of multi-biometric systems
Chiara Design and Development of multi-biometric systems
Jagtap Dipali et al. IRIS Biometrics for Embedded Systems

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: 20774228

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2021555011

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 20217033357

Country of ref document: KR

Kind code of ref document: A

ENP Entry into the national phase

Ref document number: 2020774228

Country of ref document: EP

Effective date: 20211015