Device and method of classifying fingerprint pattern type using deep learning

Abstract

A method for classifying fingerprint shape types using deep learning according to an embodiment of the present invention may include: acquiring a plurality of fingerprint images; classifying and labeling the acquired fingerprint images based on a predefined classification system; creating a fingerprint image database by storing the classified and labeled fingerprint images; extracting training data and validation data from the fingerprint image database and separating them according to a pre-set ratio; training a deep learning model using the training data; verifying the trained deep learning model using the validation data; classifying the shape of the fingerprint images using the trained deep learning model; and performing detailed classification for each shape through machine learning with the result of the fingerprint image shape classification as input.

Description

Technology Field

[0001] Embodiments of the present invention relate to a fingerprint shape type classification device and method using deep learning. Background Technology

[0002] Fingerprint evidence found at a crime scene is one of the important pieces of evidence for identifying the culprit.

[0003] Once fingerprints are secured at a crime scene, the process of identifying their owners is undertaken. Currently, the National Police Agency uses the Automatic Fingerprint Identification System (AFIS) to obtain a list of key suspects. Once the list is secured, experts determine the identity of the fingerprints through a process consisting of analysis, comparison, evaluation, and verification.

[0004] During the analysis, the pattern type of the fingerprint is classified, and the quality level of the fingerprint is evaluated. Classifying the pattern type allows for the exclusion and screening out of fingerprints with different pattern types.

[0005] When there are many fingerprints to compare, analysis requires significant time and manpower. Therefore, an automated fingerprint classification system is necessary because automating the process can reduce time and manpower, and also enhance the reliability of the classification through cross-verification.

[0006] Currently, significant technological advancements are being made in the field of forensic science using deep learning, and there is a need to apply deep learning to the field of fingerprint classification as well. The problem to be solved

[0007] Embodiments of the present invention aim to provide a fingerprint shape type classification device and method using deep learning, which reduces expert fatigue during field analysis, improves analysis speed, and improves accuracy through automatic classification of fingerprints. means of solving the problem

[0008] A method for classifying fingerprint shape types using deep learning according to an embodiment of the present invention may include: acquiring a plurality of fingerprint images; classifying and labeling the acquired fingerprint images based on a predefined classification system; creating a fingerprint image database by storing the classified and labeled fingerprint images; extracting training data and validation data from the fingerprint image database and separating them according to a pre-set ratio; training a deep learning model using the training data; verifying the trained deep learning model using the validation data; classifying the shape of the fingerprint images using the trained deep learning model; and performing detailed classification for each shape through machine learning with the result of the fingerprint image shape classification as input.

[0009] In one embodiment of the present invention, the fingerprint image may include an original fingerprint image and an extended fingerprint image.

[0010] In one embodiment of the present invention, the extended fingerprint image can be generated by dividing the original fingerprint image into a plurality of regions and applying an opaque filter to the divided fingerprint images.

[0011] In one embodiment of the present invention, the shape classification of the fingerprint image and the detailed classification results for each shape can be output.

[0012] In one embodiment of the present invention, based on the verification result of the trained deep learning model, additional training of the trained deep learning model can be performed using additional fingerprint images.

[0013] In one embodiment of the present invention, the shape classification of the fingerprint image may include a first classification of the fingerprint image into loops, whorls, arcs, and other patterns according to the shape characteristics of the ridges, and a second classification in which the loops are classified into right-sloping loops and left-sloping loops, the whorls are classified into flat whorls, double-loop whorls, central pocket whorls, and right whorls, the flat whorls are classified into genuine whorls and annular whorls, the genuine whorls and annular whorls are classified into clockwise rotation and counterclockwise rotation, the double-loop whorls are classified into clockwise rotation and counterclockwise rotation, the arcs are classified into flat arcs and protruding arcs, and the other patterns are classified into multiple layers or extended scars and surgical cut marks.

[0014] In one embodiment of the present invention, detailed classification of each shape of the fingerprint image can be performed by verifying the number of ridges and the location of the trace lines.

[0015] In one embodiment of the present invention, the detailed classification for each shape of the fingerprint image is such that the right-sloping pattern and the left-sloping pattern are classified according to the number of ridges crossing an imaginary line between the center and the triangle, and the clockwise rotation and counterclockwise rotation of the clear whorl pattern, the annular pattern, and the double ring whorl pattern can be classified into internal tracking, external tracking, and meeting tracking.

[0016] A fingerprint shape type classification device using deep learning according to one embodiment of the present invention includes at least one memory for storing at least one instruction and at least one processor, wherein the at least one processor executes the at least one instruction to acquire a plurality of fingerprint images, classify and label the acquired fingerprint images based on a predefined classification system, store the classified and labeled fingerprint images to create a fingerprint image database, extract training data and validation data from the fingerprint image database, separate them according to a pre-set ratio, train a deep learning model using the training data, validate the trained deep learning model using the validation data, classify the shape of the fingerprint images using the trained deep learning model, and perform detailed classification for each shape through machine learning using the result of the fingerprint image shape classification as input.

[0017] In one embodiment of the present invention, the at least one processor can output a shape classification of the fingerprint image and a detailed classification result for each shape by executing the at least one instruction.

[0018] In one embodiment of the present invention, the at least one processor can perform additional training of the trained deep learning model using additional fingerprint images based on the verification result of the trained deep learning model by executing the at least one instruction.

[0019] A computer program stored in a computer-readable storage medium according to one embodiment of the present invention can execute a fingerprint shape type classification method using deep learning by using a computer.

[0020] Other aspects, features, and advantages other than those described above will become clear from the following drawings, claims, and description of the invention. Effects of the invention

[0021] According to an embodiment of the present invention, classification can be easily performed during on-site fingerprint observation, which can contribute to reducing expert fatigue during on-site analysis, increasing analysis speed, and improving accuracy.

[0022] According to an embodiment of the present invention, the accuracy of fingerprint image classification can be improved by first classifying the overall shape of the fingerprint image and then performing detailed classification related to feature points for each shape. Brief explanation of the drawing

[0023] FIG. 1 is a schematic diagram illustrating a fingerprint shape type classification device using deep learning according to one embodiment of the present invention. Figure 2 is a diagram illustrating a fingerprint classification system. FIG. 3 is a flowchart illustrating a fingerprint shape type classification method using deep learning according to an embodiment of the present invention. Figure 4 is an example diagram showing a plurality of divided regions in an original fingerprint image for generating an expanded fingerprint image. Figure 5 is an example of applying an opaque filter to a fingerprint image. Figure 6 is an example of applying one-way and two-way gradient opacity filters to a fingerprint image. FIG. 7 is a drawing illustrating the output of a display according to one embodiment of the present invention. Specific details for implementing the invention

[0024] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings. When describing with reference to the drawings, identical or corresponding components are given the same reference numerals, and redundant descriptions thereof will be omitted.

[0025] In the following embodiments, terms such as first, second, etc. are used not in a limiting sense, but for the purpose of distinguishing one component from another component.

[0026] In the following examples, singular expressions include plural expressions unless the context clearly indicates otherwise.

[0027] In the following embodiments, terms such as 'include' or 'have' mean that the features or components described in the specification are present, and do not preclude the possibility that one or more other features or components may be added.

[0028] In this specification, 'fingerprint' is a unique pattern on the fingertips of a person, and is a ridge pattern on the surface of the skin. 'Fingerprint' may largely consist of ridges, which are raised lines of sweat gland openings on the skin at the fingertips, and valleys, which are indented parts between the ridges.

[0029] FIG. 1 is a schematic diagram illustrating a fingerprint shape type classification device using deep learning according to one embodiment of the present invention.

[0030] Referring to FIG. 1, the fingerprint type classification device (10) may include a communication unit (110), a processor (120), a memory (130), a user interface (140), and a display (150). Here, the fingerprint type classification device (10) to which the present invention is applied may be an information processing device used by a user. However, the present invention is not limited thereto, and the fingerprint type classification device (10) may include additional other components, and some components may be omitted.

[0031] The fingerprint type classification device (10) may include at least one of a PC (personal computer), a laptop computer, a mobile phone, a tablet PC, a smartphone, a PDA (personal digital assistants), and a PMP (portable multimedia player).

[0032] The communication unit (110) is connected to the processor (120) and memory (130) to transmit and receive data. The communication unit (110) may be connected to other external devices to transmit and receive data. In the following, the expression "transmit and receive A" may indicate transmitting and receiving "information or data representing A".

[0033] The communication unit (110) may be implemented as a circuitry within the fingerprint type classification device (10). For example, the communication unit (110) may include an internal bus and an external bus. As another example, the communication unit (110) may be an element connecting the fingerprint type classification device (10) and an external device. The communication unit (110) may be an interface. The communication unit (110) may receive data from an external device and transmit the data to the processor (120) and memory (130). For example, the communication unit (110) may receive a fingerprint image from a capturing device.

[0034] The processor (120) can control the overall operation of the fingerprint shape type classification device (10) according to one embodiment of the present invention and perform logical operations.

[0035] The processor (120) processes data received by the communication unit (110) and data stored in memory (130). The processor (120) may be a data processing device implemented in hardware having a circuit having a physical structure for executing desired operations. For example, the desired operations may include code or instructions included in a program.

[0036] The processor (120) executes computer-readable code (e.g., software) stored in memory (e.g., memory (130)) and instructions triggered by the processor (120).

[0037] The processor (120) executes at least one instruction stored in memory (130) to control the performance of a desired operation. Here, the at least one instruction may be stored in internal memory included within the processor (120) or in memory (130) included in a data processing device separate from the processor (120).

[0038] For example, the processor (120) can perform the following operations 210 to 290, and specific details will be described later.

[0039] The memory (130) stores data received by the communication unit (110) and data processed by the processor (120). The memory (130) can store a program (or application, software) that operates the fingerprint type classification device (10). The stored program can be coded to control the fingerprint type classification device (10) and can be executed by the processor (120).

[0040] The memory (130) may include volatile memory such as SRAM (Static RAM), DRAM (Dynamic RAM), SDRAM (Synchronous DRAM), or non-volatile memory such as flash memory, PRAM (Phase-change RAM), MRAM (Magnetic RAM), RRAM (Resistive RAM), FRAM (Ferroelectric RAM).

[0041] The user interface (140) can receive user input for controlling the fingerprint type classification device (10). The user interface (140) may include, but is not limited to, a touch panel that detects a user's touch, a button that receives a user's push operation, a mouse or keyboard for designating or selecting a point on the user interface screen.

[0042] The user interface (140) may be at least one GUI (Graphical User Interface) provided to control, for example, a fingerprint type classification device (10).

[0043] The display (150) can display and output information processed by the fingerprint type classification device (10). For example, the display (150) can display a GUI that controls the fingerprint type classification device (10).

[0044] The display (150) can be implemented as an LCD, PDP, OLED, FED, LED, flexible display, and 3D display, but is not limited thereto. Additionally, the display (150) can be configured as a touch screen and used as an input device in addition to an output device.

[0045] Additionally, in other embodiments, the fingerprint shape type classification device (10) may include more components than those of FIG. 1. For example, the fingerprint shape type classification device (10) may include additional components such as a battery and charging device that power the internal components, a database, etc.

[0046] FIG. 2 is a diagram illustrating a fingerprint classification system. FIG. 3 is a flowchart illustrating a method for classifying fingerprint shape types using deep learning according to an embodiment of the present invention. FIG. 4 to 6 are diagrams illustrating the process of generating an extended fingerprint image.

[0047] Referring to FIGS. 2 to 6, a fingerprint shape type classification method (20) using deep learning according to one embodiment of the present invention will be described.

[0048] The following operations can be performed by the aforementioned fingerprint shape type classification device (10).

[0049] In operation 210, the fingerprint shape type classification device (10) can acquire multiple fingerprint images.

[0050] The fingerprint type classification device (10) can receive a plurality of fingerprint images acquired from a shooting device through a communication unit (110). Additionally, the fingerprint type classification device (10) may include a camera including a lens and an image sensor. The image sensor can convert an image input by the lens into an electrical signal and may be a semiconductor device such as a CCD or CMOS.

[0051] Multiple fingerprint images may include an original fingerprint image and an extended fingerprint image.

[0052] Since the use of fingerprints involves the human body, deliberation and approval procedures by an Institutional Review Board (IRB) may be required. Original fingerprint images can be obtained by photographing approved fingerprints or those acquired at a crime scene using a camera.

[0053] The acquired fingerprint image may include a real-life image captured by a camera, a scanned document image, etc.

[0054] The acquired fingerprint images undergo a preprocessing stage. This preprocessing may include resizing, noise removal, brightness / contrast normalization, and rotation / inversion correction. This is intended to improve the accuracy of subsequent classification tasks and ensure data consistency.

[0055] In addition, for more accurate classification, an extended fingerprint image can be generated for each original fingerprint image. An extended fingerprint image can be generated through various transformations and processing based on the acquired original fingerprint image. The extended fingerprint image can be generated by methods such as rotation, resizing, color conversion, and the addition of noise to the original fingerprint image.

[0056] According to one embodiment of the present invention, an extended fingerprint image can be generated by dividing an original fingerprint image into a plurality of regions and applying an opaque filter to the divided fingerprint images.

[0057] Referring to Fig. 4, for example, each original fingerprint image is divided into 10 equal parts horizontally and vertically to form a total of 100 regions. The divided regions can be combined to include at least two sections horizontally and vertically, and stored in a database in the form of a rectangle or a square.

[0058] Referring to FIG. 5, an overall opacity filter can be applied to the segmented fingerprint image in increments of 10% in the range of 0% to 90%. FIG. 5A, FIG. 5B, and FIG. 5C show the segmented fingerprint image with an overall opacity filter applied at 0%, 50%, and 90%, respectively.

[0059] Referring to FIG. 6, a unidirectional gradient opacity filter can be applied to a segmented fingerprint image by rotating it in increments of 0 to 45 degrees (Fig. 6A). Additionally, a bidirectional gradient opacity filter can be applied to a segmented fingerprint image by setting the center to be clearly visible and then rotating it in increments of 0 to 45 degrees (Fig. 6B).

[0060] In operation 220, the acquired fingerprint image can be classified and labeled based on a predefined classification system.

[0061] Here, the predefined classification system may be a fingerprint classification system classified according to the morphological characteristics of the fingerprint. The fingerprint classification system may include first to third classifications.

[0062] The first classification allows fingerprints to be classified into loop, whorl, arch, and other patterns according to the morphological characteristics of the ridges.

[0063] The second category is a subcategory of the first category and is as follows.

[0064] Defrost doors can be classified into right-sloping defrost doors (Right loop) and left-sloping defrost doors (Left loop).

[0065] Whorls can be classified into plain whorls, double loop whorls, central pocket loop whorls, and accidental whorls. Plain whorls are classified into regular whorls and annular whorls, and regular whorls and annular whorls can be classified into clockwise rotation and counterclockwise rotation. Double loop whorls can be classified into clockwise rotation and counterclockwise rotation.

[0066] Arch patterns can be classified into plain arches and tented arches.

[0067] Other scars can be classified into mutilated or extensively scarred scars and amputation scars.

[0068] The third classification is a sub-classification of the second classification. In addition, the third classification is a detailed classification for each type of fingerprint, and can be classified by checking the number of ridges and the location of trace lines for each type.

[0069] For example, the detailed classification for each type of fingerprint is as follows.

[0070] Right-sloping and left-sloping patterns can be classified according to the number of ridges crossing the imaginary line between the center and the triangle.

[0071] Clockwise and counterclockwise rotations of pure vortex patterns, annular patterns, and double-ring vortex patterns can be classified into inner tracing, outer tracing, and meeting tracing.

[0072] Labeling can be performed on the fingerprint images classified as above.

[0073] Labeling is the process of identifying the content and features of an image and tagging them in a form suitable for machine learning.

[0074] Operation 230 can create a fingerprint image database by storing classified and labeled fingerprint images.

[0075] Classified and labeled fingerprint images can be stored in a hierarchical structure according to the assigned classification system and labels. Each fingerprint image is assigned a unique identifier, and metadata may be recorded along with it. For example, metadata may include classification information and labels of the fingerprint image, information distinguishing between original and extended fingerprint images, preprocessing history, transformation information used for data expansion, and characteristic information of the fingerprint image (size, format, etc.). Metadata facilitates the search, filtering, and management of the database and enables efficient data selection and utilization during the training process of deep learning models.

[0076] In addition, the fingerprint image database can be designed with an expandable structure to facilitate the addition of new fingerprint images or classification systems.

[0077] Operation 240 can extract training data and validation data from a fingerprint image database and separate them according to a preset ratio.

[0078] A sufficient amount of data is required to train deep learning and machine learning models, and this data is divided into training data and validation data. Training data is used for the model to learn patterns, while validation data is used to evaluate the performance of the trained model.

[0079] Extract data suitable for training from the fingerprint image database.

[0080] Afterwards, the data is separated into training data and validation data according to a pre-set ratio (e.g., 80:20 or 70:30, etc.).

[0081] This process can be performed automatically without manual intervention and can be applied based on random sampling of data or specific criteria (e.g., maintaining class balance). These data splitting methods can contribute to improving the accuracy of learning and preventing overfitting.

[0082] In operation 250, deep learning model training can be performed using training data.

[0083] Deep learning is a field of machine learning that utilizes multilayer neural networks to learn data and automatically extract features to recognize patterns. For example, deep learning models applicable to the embodiments of the present invention include CNN (Convolutional Neural Network) and ResNet (Residual Network).

[0084] Training data can be input into a deep learning model, and weights can be optimized using a loss function and an optimizer.

[0085] The loss function can quantify and measure the difference (loss) between the actual and predicted values ​​of a deep learning model. The weights and biases of a deep learning model are updated based on the loss function, and optimizers (e.g., Adam, SGD) can optimize the model using gradient descent.

[0086] Optimal performance can be achieved by adjusting hyperparameters such as the learning rate, batch size, and number of epochs during the training process of a deep learning model.

[0087] In operation 260, verification of the trained deep learning model can be performed using the verification data.

[0088] The performance of the model can be evaluated using validation data that was not used as training data. During the validation process, evaluation metrics can be calculated by comparing the predicted values ​​output by the deep learning model with the actual correct values. These evaluation metrics may include accuracy, precision, recall, F1-score, and loss function values.

[0089] For example, the deep learning model uses ResNet-101, and training is performed over 100,000 epochs. The learning rate is set to 0.001, and validation can be performed in 10 cycles.

[0090] After the operation (260) of performing verification of the trained deep learning model, the next operation 265 may be added.

[0091] In operation 265, based on the validation results of the trained deep learning model, additional training of the trained deep learning model can be performed using additional fingerprint images.

[0092] Here, the additional fingerprint images are the fingerprint images that were not used for training and validation.

[0093] In the additional training process, the performance of the model can be optimized by retraining the existing deep learning model using additional fingerprint images. To this end, the additional fingerprint images can be used as training data after undergoing a preprocessing process to improve data quality.

[0094] In addition, according to one embodiment of the present invention, instead of retraining all weights of an existing deep learning model, a fine-tuning method can be applied to adjust only the weights for a specific layer of the deep learning model.

[0095] In operation 270, shape classification of fingerprint images can be performed using a trained deep learning model.

[0096] Using a trained deep learning model, key features can be extracted from an input fingerprint image, and the overall shape of the fingerprint image can be classified based on the key features.

[0097] The classification of fingerprint image shapes may include the first and second classifications described above.

[0098] The first classification can classify fingerprint images into loop, whorl, arch, and other patterns according to the morphological characteristics of the ridges.

[0099] The second category is a subcategory of the first category.

[0100] Defrost doors can be classified into right-sloping defrost doors (Right loop) and left-sloping defrost doors (Left loop).

[0101] Whorls can be classified into plain whorls, double loop whorls, central pocket loop whorls, and accidental whorls. Plain whorls are classified into regular whorls and annular whorls, and regular whorls and annular whorls can be classified into clockwise rotation and counterclockwise rotation. Double loop whorls can be classified into clockwise rotation and counterclockwise rotation.

[0102] Arch patterns can be classified into plain arches and tented arches.

[0103] Other scars can be classified into mutilated or extensively scarred scars and amputation scars.

[0104] In operation 280, detailed classification for each shape can be performed through machine learning using the shape classification results of the fingerprint image as input.

[0105] Detailed classification can be performed using machine learning with the aforementioned shape classification results as input. For example, machine learning models such as Random Forest, Support Vector Machine (SVM), or neural networks with more complex structures can be applied.

[0106] For example, detailed classification of each type of fingerprint image is done by identifying the number of ridges and the location of trace lines.

[0107] Specifically, right-sloping and left-sloping vortices can be classified according to the number of ridges crossing the imaginary line between the center and the triangle. Clockwise and counterclockwise rotations of clear vortices, annular vortices, and double-ring vortices can be classified into inner tracing, outer tracing, and meeting tracing.

[0108] In operation 290, the fingerprint shape type classification device (10) can output a classification result.

[0109] The operation of outputting the classification result (290) may include the operation of classifying the shape of the fingerprint image and displaying the detailed classification result for each shape (e.g., the result of the first to third classifications).

[0110] FIG. 7 is a drawing illustrating the output of a display according to an embodiment of the present invention.

[0111] For example, referring to FIG. 7, the classification results for a specific fingerprint image can be displayed on the display (150). When a user selects a fingerprint image through the 'Image' button (151) and classifies it through the 'Classification' button (153), it can be displayed as 'Whisker pattern - Flat whorl pattern - Circular whorl pattern - Clockwise rotation - Internal tracking'.

[0112] As another example, the GUI may be used to highlight nodes corresponding to specific classification results on the logic tree. Node highlighting can be achieved by changing the node's color or modifying its configuration (node ​​shape, font weight, etc.). Highlighting can also be achieved by expanding or contracting the node's size. Thus, the user can check the location of the corresponding classification result on the logic tree.

[0113] Meanwhile, the fingerprint shape type classification method (20) according to various embodiments of the present invention illustrated in FIG. 3 can be written as a program that can be executed on a computer and can be implemented on a general-purpose digital computer that operates the program using a computer-readable recording medium. The computer-readable recording medium may include storage media such as magnetic storage media (e.g., ROM, floppy disk, hard disk, etc.) and optical reading media (e.g., CD-ROM, DVD, etc.).

[0114] The above description is merely an illustrative explanation of the technical concept of the present invention, and those skilled in the art to which the present invention pertains will be able to make various modifications and variations within the scope of the essential characteristics of the present invention. For example, appropriate results may be achieved even if the described techniques are performed in a different order than the described method, and / or the components of the described system, structure, device, circuit, etc. are combined or assembled in a form different from the described method, or are replaced or substituted by other components or equivalents.

[0115] Accordingly, the scope of protection of the present invention shall be interpreted by the claims below, and all technical ideas within an equivalent scope shall be interpreted as being included within the scope of rights of the present invention. Explanation of the symbols

[0116] 10: Fingerprint shape type classification device 110 : Communications Department 120 : Processor 130 : Memory 140 : User Interface 150 : Display

Claims

Claim 1 A method for classifying fingerprint shape types using deep learning, comprising: acquiring a plurality of fingerprint images; classifying and labeling the acquired fingerprint images based on a predefined classification system; creating a fingerprint image database by storing the classified and labeled fingerprint images; extracting training data and validation data from the fingerprint image database and separating them according to a preset ratio; training a deep learning model using the training data; verifying the trained deep learning model using the validation data; classifying the shape of the fingerprint images using the trained deep learning model; and performing detailed classification for each shape through machine learning using the result of the shape classification of the fingerprint images as input. Claim 2 A method for classifying fingerprint shape types using deep learning, wherein the fingerprint image includes an original fingerprint image and an extended fingerprint image in claim 1. Claim 3 A method for classifying fingerprint shape types using deep learning, wherein, in paragraph 2, the expanded fingerprint image is generated by dividing the original fingerprint image into multiple regions and applying an opaque filter to the divided fingerprint images. Claim 4 A fingerprint shape type classification method using deep learning, further comprising, in claim 1, the operation of classifying the shape of the fingerprint image and outputting a detailed classification result for each shape. Claim 5 A method for classifying fingerprint shape types using deep learning, further comprising, in claim 1, the operation of performing additional training of the trained deep learning model using additional fingerprint images based on the verification result of the trained deep learning model. Claim 6 A method for classifying fingerprint shape types using deep learning, wherein the shape classification of the fingerprint image comprises: a first classification classifying the fingerprint image into loops, whorls, arcs, and other patterns according to the shape characteristics of the ridges; and a second classification classifying the loops into right-slanted loops and left-slanted loops, the whorls into flat whorls, double-loop whorls, central pocket whorls, and right whorls, the flat whorls into genuine whorls and annular whorls, the genuine whorls and annular whorls into clockwise rotation and counterclockwise rotation, the double-loop whorls into clockwise rotation and counterclockwise rotation, the arcs into flat arcs and protruding arcs, and the other patterns into multiple layers or extended scars and surgical cut marks. Claim 7 In claim 1, the detailed classification of each shape of the fingerprint image is a method for classifying fingerprint shape types using deep learning, wherein the classification is performed by verifying the number of ridges and the location of trace lines. Claim 8 A method for classifying fingerprint shape types using deep learning, wherein, in claim 1, the detailed classification for each shape of the fingerprint image is such that the right-sloping serrated pattern and the left-sloping serrated pattern are classified according to the number of ridges crossing an imaginary line between the center and the triangle, and the clockwise and counterclockwise rotations of the clear whorl pattern, the annular pattern, and the double ring whorl pattern are classified into internal tracking, external tracking, and intersection tracking. Claim 9 A fingerprint shape type classification device using deep learning, comprising: at least one memory storing at least one instruction; and at least one processor; wherein the at least one processor, by executing the at least one instruction, acquires a plurality of fingerprint images, classifies and labels the acquired fingerprint images based on a predefined classification system, stores the classified and labeled fingerprint images to create a fingerprint image database, extracts training data and validation data from the fingerprint image database, separates them according to a preset ratio, performs training of a deep learning model using the training data, performs validation of the trained deep learning model using the validation data, performs shape classification of the fingerprint images using the trained deep learning model, and performs detailed classification for each shape through machine learning using the result of shape classification of the fingerprint images as input. Claim 10 In claim 9, the fingerprint image comprises an original fingerprint image and an extended fingerprint image, a fingerprint shape type classification device using deep learning. Claim 11 A fingerprint shape type classification device using deep learning, wherein the expanded fingerprint image is generated by dividing the original fingerprint image into multiple regions and applying an opaque filter to the divided images. Claim 12 A fingerprint shape type classification device using deep learning, wherein, in claim 9, the at least one processor outputs a shape classification of the fingerprint image and a detailed classification result for each shape by executing the at least one instruction. Claim 13 A fingerprint shape type classification device using deep learning, wherein, in claim 9, the at least one processor performs additional training of the trained deep learning model using additional fingerprint images based on the verification result of the trained deep learning model by executing the at least one instruction. Claim 14 A fingerprint shape type classification device using deep learning, wherein the shape classification of the fingerprint image comprises: a first classification classifying the fingerprint image into loops, whorls, arcs, and other patterns according to the shape characteristics of the ridges; and a second classification classifying the loops into right-slanted loops and left-slanted loops, the whorls into flat whorls, double-loop whorls, central pocket whorls, and right whorls, the flat whorls into genuine whorls and annular whorls, the genuine whorls and annular whorls into clockwise rotation and counterclockwise rotation, the double-loop whorls into clockwise rotation and counterclockwise rotation, the arcs into flat arcs and protruding arcs, and the other patterns into multiple layers or extended scars and surgical cut marks. Claim 15 In claim 9, the detailed classification for each form of the fingerprint image is performed by a deep learning fingerprint form type classification device that classifies by verifying the number of ridges and the location of trace lines. Claim 16 A fingerprint shape type classification device using deep learning, wherein, in claim 9, the detailed classification for each shape of the fingerprint image is such that the right-sloping serrated pattern and the left-sloping serrated pattern are classified according to the number of ridges crossing an imaginary line between the center and the triangle, and the clockwise and counterclockwise rotations of the clear whorl pattern, the annular pattern, and the double ring whorl pattern are classified into internal tracking, external tracking, and intersection tracking. Claim 17 A computer program stored on a computer-readable storage medium to execute the method of any one of claims 1 through 8 using a computer.

Images