CN116071786A - Fingerprint identification device and method thereof - Google Patents

Fingerprint identification device and method thereof Download PDF

Info

Publication number
CN116071786A
CN116071786A CN202310002883.9A CN202310002883A CN116071786A CN 116071786 A CN116071786 A CN 116071786A CN 202310002883 A CN202310002883 A CN 202310002883A CN 116071786 A CN116071786 A CN 116071786A
Authority
CN
China
Prior art keywords
fingerprint
image
frame
true
false
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202310002883.9A
Other languages
Chinese (zh)
Inventor
冯继雄
王长海
田志民
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chipone Technology Beijing Co Ltd
Original Assignee
Chipone Technology Beijing Co Ltd
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 Chipone Technology Beijing Co Ltd filed Critical Chipone Technology Beijing Co Ltd
Priority to CN202310002883.9A priority Critical patent/CN116071786A/en
Publication of CN116071786A publication Critical patent/CN116071786A/en
Pending legal-status Critical Current

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
    • G06V40/1335—Combining adjacent partial images (e.g. slices) to create a composite input or reference pattern; Tracking a sweeping finger movement
    • 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
    • G06V40/1382—Detecting the live character of the finger, i.e. distinguishing from a fake or cadaver finger

Landscapes

  • Engineering & Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Collating Specific Patterns (AREA)

Abstract

The invention discloses a fingerprint identification device and a method thereof, wherein the device comprises the steps of continuously collecting multi-frame fingerprint images; detecting the moving distance of two adjacent frames of fingerprint images, and generating a moving complex distance matrix; normalizing the moving complex distance matrix to obtain a moving speed matrix; and the movement speed matrix is formed into an input tensor and is sent into a trained fingerprint identification model to judge whether the fingerprint is true or false, so that the safety of the fingerprint identification device is improved.

Description

Fingerprint identification device and method thereof
Technical Field
The invention relates to the technical field of fingerprint identification, in particular to a fingerprint identification device and a fingerprint identification method.
Background
The fingerprint recognition technology forms a fingerprint image through the electrical or optical difference of the fingerprint valley and ridge, and then recognizes the image, and the technology is widely applied to various mobile intelligent terminals, door locks, automobiles and other devices. In daily life, a person may leave a fingerprint trace at a location where many fingers touch. At present, false fingerprints forged by various bionic materials appear, and the appearance of the false fingerprints reduces the safety of the fingerprint identification device.
The existing fingerprint identification technology utilizes the difference between an image acquired by a fake fingerprint and an image acquired by a real finger to perform anti-counterfeiting identification, and the principle is that the conductivity/reflectivity of the fake fingerprint is different from that of a human finger, the gray level of the acquired image is different, the image acquired by the real finger is shown in fig. 1a, and the image acquired by the fake fingerprint is shown in fig. 1 b; in addition, the valley ridge line of the image acquired by the fake fingerprint is more prominent as shown in fig. 1c, and is different from the valley ridge line of the image acquired by the real finger. But these differences become smaller and smaller with material and process improvements, and are barely distinguishable from the image alone.
Therefore, a new fingerprint identification method is to be proposed to accurately judge whether the fingerprint is true or false.
Disclosure of Invention
In view of the above problems, an object of the present invention is to provide a fingerprint identification apparatus and a fingerprint identification method, so that the fingerprint can be accurately determined based on the principle that friction coefficients of different fingerprint molds on the surface of a fingerprint sensor are different.
According to an aspect of the present invention, there is provided a fingerprint identification method including continuously acquiring a plurality of frames of fingerprint images; detecting the moving distance of two adjacent frames of fingerprint images, and generating a moving complex distance matrix; normalizing the moving complex distance matrix to obtain a moving speed matrix; and forming the movement speed matrix into an input tensor, and sending the input tensor into a trained fingerprint identification model to judge whether the fingerprint is true or false.
Optionally, when the fingerprint judgment result is a true fingerprint, selecting at least one frame of fingerprint image to match with a preset fingerprint image, judging that the fingerprint identification is successful according to the matching result of successful matching, and judging that the fingerprint identification is failed according to the matching result of failed matching; and when the fingerprint judgment result is a fake fingerprint, judging that the fingerprint identification fails.
Optionally, the fingerprint identification method further includes selecting at least one frame of fingerprint image to match with a preset fingerprint image after the continuous collection of the multi-frame fingerprint images, judging fingerprint identification failure according to a matching result of the matching failure, judging whether the fingerprint is true or false according to a matching result of the matching success, judging that the fingerprint identification is successful when the fingerprint judgment result is true, and judging that the fingerprint identification is failed when the fingerprint judgment result is false.
Optionally, the continuously acquiring the multi-frame fingerprint image includes continuously scanning the finger fingerprint at preset time intervals to acquire the multi-frame fingerprint image.
Optionally, detecting the moving distance of two adjacent frames of fingerprint images, and generating a moving complex distance matrix, wherein the method comprises selecting an image with a preset size from the first fingerprint images as a first image; searching a second image which has the smallest difference with the first image and has the same size in the next frame of fingerprint image of the first fingerprint image to obtain the optimal offset vector (dx, dy) of the pixel units in the first fingerprint image and the next frame of fingerprint image, wherein the first fingerprint image is any frame of fingerprint image in the multi-frame fingerprint image except the last frame of fingerprint image; the complex z=dx+i×dy is formed according to the optimal offset vector (dx, dy) to obtain the moving complex distance matrix, B 1 ，B 2 ，……，B N-1 Wherein N represents the number of the fingerprint images and i represents the acquisition order of the first fingerprint image.
Optionally, the movement velocity matrix C i ＝B i /(t i+1 -t i ) Wherein B represents a moving complex distance matrix, t i+1 -t i Representing the interval time of collecting fingerprint images of adjacent frames, and i represents the collection sequence of the first fingerprint images.
Optionally, the first image is obtained by selecting an image with a preset size from the center of the first fingerprint image, and the second image is obtained by searching an image with the smallest difference and the same size as the first image within a preset radius by taking the center of the fingerprint image of the next frame of the first fingerprint image as the center of the circle.
Optionally, the fingerprint identification model is a convolutional neural network model, and the training method comprises the following steps: acquiring a true and false fingerprint training sample set; inputting the moving speed matrix in the true and false fingerprint training sample set into a convolutional neural network for deep learning training to obtain the convolutional neural network model, wherein the true and false fingerprint training sample set comprises: a real fingerprint training sample set, a moving speed matrix set obtained after a plurality of different fingers press the fingerprint sensor; a false fingerprint training sample set, and a moving speed matrix set obtained after a fingerprint mould made of a plurality of different materials presses the fingerprint sensor.
According to another aspect of the present invention, there is provided a fingerprint identification apparatus, including an acquisition module including an acquisition unit configured to continuously acquire a plurality of frames of fingerprint images and a calculation unit configured to detect a moving distance of two adjacent frames of fingerprint images and acquire a moving speed matrix according to the moving distance; and the fingerprint identification model is set to judge whether the fingerprint is true or false according to the moving speed matrix.
Optionally, the fingerprint identification device further comprises a matching module, wherein the matching module is configured to receive a fingerprint judgment result of the true fingerprint output by the fingerprint identification model, and select at least one frame of fingerprint image to match with a preset fingerprint image; and the judging module is used for outputting a fingerprint identification result according to the matching result and the fingerprint judging result.
Optionally, the fingerprint identification device further comprises: the matching module is further configured to select at least one frame of fingerprint image to match with a preset fingerprint image, and output a matching result of successful matching to the acquisition module, so that the acquisition module acquires the moving speed matrix and outputs the moving speed matrix to the fingerprint identification model to judge whether the fingerprint is true or false; and the judging module is used for outputting a fingerprint identification result according to the matching result and the fingerprint judging result.
Optionally, the acquisition unit scans the finger fingerprint through the fingerprint sensor according to a preset time interval to acquire the multi-frame fingerprint image.
Optionally, the fingerprint identification model is a convolutional neural network model, the convolutional neural network model is obtained by inputting a true and false fingerprint training sample set into a convolutional neural network for deep learning training, and the true and false fingerprint training sample set includes: a real fingerprint training sample set, a moving speed matrix set obtained after a plurality of different fingers press the fingerprint sensor; a false fingerprint training sample set is a moving speed matrix set obtained after a fingerprint sensor is pressed by a fingerprint mould made of a plurality of different materials.
According to the fingerprint identification device and the fingerprint identification method, provided by the invention, the principle that the friction coefficients of fingerprint molds made of different materials on the surface of the fingerprint sensor are different and the moving speed matrix is also different is utilized, the moving distances of two adjacent frames of fingerprint images are detected by continuously collecting multi-frame fingerprint images, the moving speed matrix is obtained by normalizing the moving distances, the moving speed matrix is formed into an input tensor and is sent to the fingerprint identification model to judge whether the fingerprint is true or false, so that the fingerprint is true or false, the safety of the fingerprint identification device is improved, and the hardware cost is not increased.
Drawings
The above and other objects, features and advantages of the present invention will become more apparent from the following description of embodiments of the present invention with reference to the accompanying drawings, in which:
FIG. 1 shows a comparison of images acquired by live and fake fingerprints;
FIG. 2 shows a method flow diagram of a fingerprint identification method according to an embodiment of the invention;
fig. 3 is a schematic view showing the structure of a fingerprint recognition device according to a first embodiment of the present invention;
fig. 4 shows a schematic structural view of a fingerprint recognition device according to a second embodiment of the present invention.
Fig. 5 shows a method flow chart of a method for acquiring a fingerprint identification model according to an embodiment of the invention.
Detailed Description
Various embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. The same elements or modules are denoted by the same or similar reference numerals in the various figures. For clarity, the various features of the drawings are not drawn to scale.
It should be appreciated that in the following description, a "circuit" may include a single or multiple combined hardware circuits, programmable circuits, state machine circuits, and/or elements capable of storing instructions for execution by the programmable circuits. When an element or circuit is referred to as being "connected to" another element or circuit "connected between" two nodes, it may be directly coupled or connected to another element or intervening elements may be present, the connection between elements may be physical, logical, or a combination thereof. In contrast, when an element is referred to as being "directly coupled to" or "directly connected to" another element, it means that there are no intervening elements present between the two.
Also, certain terms are used throughout the description and claims to refer to particular components. It will be appreciated by those of ordinary skill in the art that a hardware manufacturer may refer to the same component by different names. The present patent specification and claims do not take the form of an element or components as a functional element or components as a rule.
Furthermore, it should be noted that relational terms such as first and second are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The inventor of the present application finds that the friction coefficients of fingerprint molds made of different materials on the surface of a fingerprint sensor are different, and the moving speed of a plurality of frames of fingerprint images collected when the fingerprint mold is used for pressing the fingerprint sensor is different under the influence of the fingerprint mold making materials.
Fig. 2 shows a method flowchart of a fingerprint identification method according to an embodiment of the invention, comprising the steps of:
step S101: and continuously collecting a plurality of frames of fingerprint images.
In step S101, the fingerprint image is an image of a finger fingerprint, and the method for capturing the fingerprint image involves various methods, for example, the finger fingerprint may be continuously scanned during the process of pressing the fingerprint sensor by the user to capture a plurality of frames of fingerprint images a 1 ，A 2 ，……，A N The acquisition time is t1, t2, … … and tN respectively, the acquisition time intervals of two adjacent frames of fingerprint images can be set to be the same preset time, and the acquisition time intervals of two adjacent frames of fingerprint images can also be set as required, so that each frame of fingerprint image is different due to slight movement in the finger pressing process, wherein N represents the number of frames of the fingerprint images and is a positive integer not less than 2.
Step S102: and detecting the moving distance of two adjacent frames of fingerprint images.
In step S102, at A i Randomly selecting an image with a preset size from the frame fingerprint image as a first image, wherein the first image is formed by m x n pixel units, and then at A i+1 Searching a second image with the smallest difference with the first image in the frame fingerprint image, wherein the second image is composed of m-n pixel units, and obtaining A according to the first image and the second image i Frame fingerprint image and A i+1 Optimal offset vectors (dx, dy) of pixel units in the fingerprint image of the frame form complex numbers z=dx+i×dy, and a moving complex number distance matrix of each pixel unit is obtained, B 1 ，B 2 ，……B N-1 Wherein i=1, 2,3 … …, N-1. Preferably, it can be as follows i Selecting an image with a preset size as a first image from the center of the frame fingerprint image, and then taking A as a first image i+1 Searching a second image with the smallest difference with the first image and the same size in a radius r by taking the center of the frame fingerprint image as the center of the circle, wherein r is a preset value, so that the frame fingerprint image can be saved in A i+1 The time of searching for the second image in the frame fingerprint image.
Step S103: and acquiring a moving speed matrix of the two adjacent frames of fingerprint images according to the moving distance of the two adjacent frames of fingerprint images.
In step S103, the moving complex distance matrix of two adjacent frames of fingerprint images is normalized, B 1 ，B 2 ，……B N-1 Obtaining a moving speed matrix C i ＝B i /(t i+1 -t i ) Wherein i=1, 2,3, … …, N-1, t i+1 -t i Representing the time interval between acquisition of fingerprint images of adjacent frames.
Step S104: and forming an input tensor by the movement speed matrix, and sending the input tensor into a trained fingerprint identification model to judge whether the fingerprint of the finger is a true fingerprint or not.
In step S104, the fingerprint recognition model is a convolutional neural network model, and the convolutional neural network model is obtained by performing deep learning training on a convolutional neural network by using a true and false fingerprint training sample set. Of course, the invention is not limited thereto, and in other embodiments, the set of true and false fingerprint training samples may be used to train other deep learning networks to form a fingerprint recognition model. The true and false fingerprint training sample set contains a large amount of true and false fingerprint data.
Step S105: if the fingerprint is a true fingerprint, selecting at least one frame of fingerprint image to compare with a preset fingerprint image, and judging whether the fingerprint images are matched.
In step S105, the number of the preset fingerprint images may be one or more, and in order to improve the accuracy of fingerprint identification, a plurality of frames of fingerprint images may be selected for comparison with the preset fingerprint images.
Step S106: if the selected fingerprint image is matched with the preset fingerprint image, the fingerprint identification is successful.
Step S107: if the selected fingerprint image is not matched with the preset fingerprint image, fingerprint identification fails.
Step S108: if the finger fingerprint is a fake fingerprint, the fingerprint identification fails.
In another embodiment, at least one frame of fingerprint image can be selected and compared with a preset fingerprint image, if the fingerprint image is determined to be matched with the preset fingerprint image, the moving distance of two adjacent frames of fingerprint images is detected, and the moving speed matrix is determined so as to judge whether the fingerprint is true or false and output a fingerprint identification result; if the fingerprint image is not matched with the preset fingerprint image, the fingerprint identification failure is directly determined.
Fig. 3 shows a schematic configuration of a fingerprint recognition device according to a first embodiment of the present invention. The fingerprint recognition device 100 includes an acquisition module 110, a fingerprint recognition model 120, a matching module 130, and a judgment module 140.
The acquisition module 110 continuously acquires a plurality of frames of fingerprint images and acquires a moving speed matrix of two adjacent frames of fingerprint images. The acquisition module 110 includes an acquisition unit 111 and a calculation unit 112.
The acquisition unit 111 continuously scans the finger fingerprint through the fingerprint sensor to obtain multi-frame fingerprint images, and the acquisition unit 111 may be configured to scan once every predetermined time interval, sort the multi-frame fingerprint images according to the acquisition sequence of the fingerprint images, and input the multi-frame fingerprint images to the calculation unit 112.
The computing unit 112 performs the fingerprint image a in the first frame 1 Randomly selecting an image with a preset size as a first image, and performing fingerprint image A on a second frame 2 Searching a second image which has the smallest difference with the first image and has the same size; then obtaining a first frame fingerprint image A according to the first image and the second image 1 And a second frame fingerprint image A 2 The optimal offset vector (dx, dy) of the middle pixel unit, which forms a complex number z=dx+i dy, obtains a second frame fingerprint image a 2 Middle pixel unit relative to first frame fingerprint image A 1 Moving complex distance matrix B of middle pixel unit 1 And so on until the fingerprint image A of the last frame is obtained N Middle pixel unit relative to penultimate frame fingerprint image A N-1 Moving complex distance matrix B of middle pixel unit N-1 Then for all the mobile complex distance matrix B 1 ，B 2 ，……B N-1 Carrying out normalization processing to obtain a moving speed matrix, C i ＝B i /(t i+1 -t i ) Wherein i=1, 2,3, … …, N-1, N represents the number of frames of the fingerprint image, is a positive integer not less than 2, t i+1 -t i Representing between acquisition of fingerprint images of adjacent framesTime interval.
The fingerprint recognition model 120 is, for example, a trained convolutional neural network model, and is used for judging whether the fingerprint is true or false according to an input tensor composed of a moving speed matrix, and inputting a fingerprint judgment result to the matching module 130 when the fingerprint is true, and inputting a fingerprint judgment result to the judging module 140 when the fingerprint is false.
Fig. 5 shows a method flowchart of a method for acquiring a fingerprint recognition model according to an embodiment of the present invention, and as shown in fig. 5, the method for acquiring a fingerprint recognition model 120 includes the steps of:
step S201, a true and false fingerprint training sample set is obtained.
The true and false fingerprint training sample set comprises a true fingerprint training sample set and a false fingerprint training sample set, wherein the true fingerprint training sample set comprises a moving speed matrix set obtained after a plurality of different fingers press the fingerprint sensor, and the false fingerprint training sample set comprises a moving speed matrix set obtained after a fingerprint mold made of a plurality of different materials presses the fingerprint sensor.
Step S202, inputting a moving speed matrix in a true and false fingerprint training sample set into a convolutional neural network for deep learning training to obtain a fingerprint identification model.
The matching module 130 selects at least one frame of fingerprint image from the acquisition unit 111 to match with a preset fingerprint image when receiving the fingerprint judgment result of the true fingerprint, and outputs the matching result to the judgment module 140.
The judging module 140 judges that the fingerprint identification fails after receiving the fingerprint judging result that the fingerprint is the fake fingerprint and the matching result that the matching fails, and judges that the fingerprint identification is successful after receiving the matching result that the matching is successful. The fingerprint identification is completed after the fingerprint is judged to be a fake fingerprint, so that the process of fingerprint matching is reduced, and the fingerprint identification device can quickly obtain a fingerprint identification result.
Fig. 4 shows a schematic structural view of a fingerprint recognition device according to a second embodiment of the present invention. The fingerprint recognition device 200 also includes an acquisition module 210, a fingerprint recognition module 220, a matching module 230, and a judgment module 240.
In this embodiment, the acquisition module 210 includes an acquisition unit 211 and a calculation unit 212, after the acquisition unit 211 acquires multi-frame fingerprint images, at least one frame of fingerprint image is sent to the matching module 230, the matching module 230 matches at least one frame of fingerprint image with a preset fingerprint image, if the matching is successful, the acquisition unit 211 in the acquisition module 210 is controlled by the matching module 230 to output the multi-frame fingerprint images to the calculation unit 212 according to the acquisition sequence so as to obtain a moving speed matrix, the fingerprint identification device 220 acquires an input tensor formed by the moving speed matrix to judge whether the fingerprint is true or false, if the fingerprint is true, the judgment module 240 judges that the fingerprint identification is successful, if the fingerprint is false, the judgment module 240 judges that the fingerprint identification is failed, and if the matching is failed, the judgment module 240 directly judges that the fingerprint identification is failed. When the fingerprint image is not matched with the preset fingerprint image, fingerprint identification is completed, the process of calculating the moving speed matrix by the calculating unit 212 is reduced, and the fingerprint identification device can quickly obtain a fingerprint identification result.
According to the fingerprint identification device and the fingerprint identification method, the multi-frame fingerprint images are continuously collected, the moving distance of two adjacent frames of fingerprint images is detected, the moving distance is normalized to obtain the moving speed matrix, the moving speed matrix is formed into the input tensor and is sent to the trained convolutional neural network model to judge whether the fingerprint is true or false, and the principle that the friction coefficients of fingerprint molds made of different materials on the surface of the fingerprint sensor are different and the moving speed matrix is also different is utilized, so that the true or false fingerprint can be accurately judged, the safety of the fingerprint identification device is improved, and the hardware cost is not increased.
Embodiments in accordance with the present invention, as described above, are not intended to be exhaustive or to limit the invention to the precise embodiments disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention and various modifications as are suited to the particular use contemplated. The scope of the invention should be determined by the appended claims and their equivalents.

Claims (13)

1. A fingerprint identification method, comprising:
continuously collecting multi-frame fingerprint images;
detecting the moving distance of two adjacent frames of fingerprint images, and generating a moving complex distance matrix;
normalizing the moving complex distance matrix to obtain a moving speed matrix;
and forming the movement speed matrix into an input tensor, and sending the input tensor into a trained fingerprint identification model to judge whether the fingerprint is true or false.
2. The fingerprint identification method of claim 1, wherein the fingerprint identification method further comprises:
when the fingerprint judgment result is a true fingerprint, selecting at least one frame of fingerprint image to match with a preset fingerprint image, judging that the fingerprint identification is successful according to the matching result of successful matching, and judging that the fingerprint identification is failed according to the matching result of failed matching;
and when the fingerprint judgment result is a fake fingerprint, judging that the fingerprint identification fails.
3. The fingerprint identification method of claim 1, wherein the fingerprint identification method further comprises:
and after the multi-frame fingerprint images are continuously collected, at least one frame of fingerprint image is firstly selected to be matched with a preset fingerprint image, fingerprint identification failure is judged according to a matching result of the matching failure, the authenticity of the fingerprint is judged according to a matching result of the matching success, fingerprint identification success is judged when the fingerprint judgment result is a true fingerprint, and fingerprint identification failure is judged when the fingerprint judgment result is a false fingerprint.
4. The fingerprint identification method of claim 1, wherein the continuously acquiring multiple frames of fingerprint images includes continuously scanning a fingerprint at preset time intervals to acquire the multiple frames of fingerprint images.
5. The fingerprint recognition method of claim 4, wherein the method of detecting the moving distance of two adjacent frames of fingerprint images and generating the moving complex distance matrix comprises:
selecting an image with a preset size from the first fingerprint image as a first image;
searching a second image which has the smallest difference with the first image and has the same size in the next frame of fingerprint image of the first fingerprint image to obtain the optimal offset vector (dx, dy) of the pixel units in the first fingerprint image and the next frame of fingerprint image, wherein the first fingerprint image is any frame of fingerprint image in the multi-frame fingerprint image except the last frame of fingerprint image;
the mobile complex distance matrix B is obtained from the complex z=dx+i×dy formed by the optimal offset vector (dx, dy) 1 ，B 2 ，……B N-1 Wherein N represents the number of the fingerprint images and i represents the acquisition order of the first fingerprint image.
6. The fingerprint recognition method according to claim 5, wherein the movement velocity matrix C i ＝B i /(t i+1 -t i ) Wherein B is i Representing a moving complex distance matrix, t i+1 -t i Representing the interval time of collecting fingerprint images of adjacent frames, and i represents the collection sequence of the first fingerprint images.
7. The fingerprint identification method according to claim 5, wherein the first image is obtained by selecting an image of a predetermined size at a center of the first fingerprint image, and the second image is obtained by searching an image of a minimum difference and an equal size from the first image within a predetermined radius of a center of a next frame of the fingerprint image of the first fingerprint image as a center of the center.
8. The fingerprint identification method according to claim 1, wherein the fingerprint identification model is a convolutional neural network model, and the training method comprises:
acquiring a true and false fingerprint training sample set;
inputting the moving speed matrix in the true and false fingerprint training sample set into a convolutional neural network for deep learning training to obtain the convolutional neural network model,
the true and false fingerprint training sample set comprises:
a real fingerprint training sample set, a moving speed matrix set obtained after a plurality of different fingers press the fingerprint sensor;
a false fingerprint training sample set, and a moving speed matrix set obtained after a fingerprint mould made of a plurality of different materials presses the fingerprint sensor.
9. A fingerprint recognition device, comprising:
the acquisition module comprises an acquisition unit and a calculation unit, wherein the acquisition unit is used for continuously acquiring multi-frame fingerprint images, and the calculation unit is used for detecting the moving distance of two adjacent frames of fingerprint images and acquiring a moving speed matrix according to the moving distance;
and the fingerprint identification model is set to judge whether the fingerprint is true or false according to the moving speed matrix.
10. The fingerprint recognition device of claim 9, wherein the fingerprint recognition device further comprises:
the matching module is used for receiving a fingerprint judgment result of the true fingerprint output by the fingerprint identification model, and selecting at least one frame of fingerprint image to match with a preset fingerprint image;
and the judging module is used for outputting a fingerprint identification result according to the matching result and the fingerprint judging result.
11. The fingerprint recognition device of claim 9, wherein the fingerprint recognition device further comprises:
the matching module is further configured to select at least one frame of fingerprint image to match with a preset fingerprint image, and output a matching result of successful matching to the acquisition module, so that the acquisition module acquires the moving speed matrix and outputs the moving speed matrix to the fingerprint identification model to judge whether the fingerprint is true or false;
and the judging module is used for outputting a fingerprint identification result according to the matching result and the fingerprint judging result.
12. The fingerprint recognition device according to claim 9, wherein the acquisition unit acquires the multi-frame fingerprint image by scanning a fingerprint of the finger at preset time intervals by a fingerprint sensor.
13. The fingerprint recognition device according to claim 9, wherein the fingerprint recognition model is a convolutional neural network model obtained by inputting a true and false fingerprint training sample set into a convolutional neural network for deep learning training, the true and false fingerprint training sample set comprising:
a real fingerprint training sample set, a moving speed matrix set obtained after a plurality of different fingers press the fingerprint sensor;
a false fingerprint training sample set is a moving speed matrix set obtained after a fingerprint sensor is pressed by a fingerprint mould made of a plurality of different materials.
CN202310002883.9A 2023-01-03 2023-01-03 Fingerprint identification device and method thereof Pending CN116071786A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310002883.9A CN116071786A (en) 2023-01-03 2023-01-03 Fingerprint identification device and method thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310002883.9A CN116071786A (en) 2023-01-03 2023-01-03 Fingerprint identification device and method thereof

Publications (1)

Publication Number Publication Date
CN116071786A true CN116071786A (en) 2023-05-05

Family

ID=86172761

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310002883.9A Pending CN116071786A (en) 2023-01-03 2023-01-03 Fingerprint identification device and method thereof

Country Status (1)

Country Link
CN (1) CN116071786A (en)

Similar Documents

Publication Publication Date Title
JP2002222424A (en) Fingerprint matching system
US20180225495A1 (en) Method for authenticating a finger of a user of an electronic device
CA2256672C (en) Biometric recognition using a classification neural network
CN109583569B (en) Multi-mode feature fusion method and device based on convolutional neural network
CN102496001B (en) Method of video monitor object automatic detection and system thereof
CN113312989B (en) Finger vein feature extraction network based on aggregated descriptors and attention
CN112766218B (en) Cross-domain pedestrian re-recognition method and device based on asymmetric combined teaching network
CN113221086B (en) Offline face authentication method and device, electronic equipment and storage medium
CN112149616A (en) Figure interaction behavior recognition method based on dynamic information
KR101601187B1 (en) Device Control Unit and Method Using User Recognition Information Based on Palm Print Image
Patil et al. Fingerprint classification using artificial neural network
CN116071786A (en) Fingerprint identification device and method thereof
CN109359616B (en) Pseudo-concatenation small-size fingerprint identification algorithm based on SIFT
Tamrakar et al. Low resolution fingerprint image verification using CNN filter and LSTM classifier
CN109345685A (en) Paper product discrimination method and device and storage medium
JP4270842B2 (en) Fingerprint verification device
KR100359920B1 (en) Discrimination method for imitative iris in iris recognition system
Daramola et al. Fingerprint verification system using support vector machine
TWI531985B (en) Palm biometric method
Deore et al. Offline signature recognition: Artificial neural network approach
JP2004246459A (en) Method of calculating rotational displacement angle between fingerprint images and fingerprint image collation device
Yong-Xia et al. A new detection method of singular points of fingerprints based on neural network
CN110705352A (en) Fingerprint image detection method based on deep learning
JPH0433065A (en) Fingerprint collating device
CN113221088B (en) User identity identification method and device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination