KR20220013882A - Method and apparatus for testing liveness using multi-phase detection sensor - Google Patents

Abstract

Disclosed are a method and device for inspecting liveness using a multi-phase detection sensor. The method for inspecting liveness comprises: a step of receiving phase images of different phases; a step of generating preprocessed phase images by performing preprocessing comprising edge reinforcement processing on the phase images; a step of generating differential images based on the preprocessed phase images; a step of generating low-resolution differential images having a lower resolution than that of the differential images; a step of generating a minimum map image based on the low-resolution differential images; and a step of performing a liveness inspection for an object shown in the phase images based on the minimum map image.

Description

Method and apparatus for testing liveness using a multi-phase detection sensor

The following disclosure relates to a liveness inspection technique using a multi-phase detection sensor.

Biometric authentication technology is a technology that authenticates a user using fingerprints, iris, voice, face, blood vessels, and the like. Biometric characteristics used for authentication differ from person to person, and there is no inconvenience of possession, the risk of theft or imitation is low, and biometric authentication technology is widely used because it has advantages that it does not change easily during a lifetime. Face authentication technology, which is one of biometric authentication technologies, is an authentication technology that determines whether a user is a legitimate user based on a face displayed in a still image or video. The face authentication technology has the advantage of being able to confirm the authentication target in a contactless manner. Recently, because of the convenience and efficiency of the face authentication technology, the face authentication technology has been widely used in various application fields such as security systems, mobile authentication, and multimedia data search.

A liveness checking method according to an embodiment includes: receiving phase images of different phases; generating preprocessed phase images by performing preprocessing including edge enhancement processing on the phase images; generating difference images based on the pre-processed phase images; generating low-resolution difference images having a lower resolution than the difference images; generating a minimum map image based on the low-resolution difference images; and performing a liveness check on the object displayed in the phase images based on the minimum map image.

The performing of the pre-processing may include removing noise from the phase images and performing processing for strengthening an edge region of an object displayed in each of the phase images.

According to an embodiment, the generating of the low-resolution difference images may include: extracting the largest pixel value for each patch region from the difference image; and generating a low-resolution difference image based on the extracted largest pixel value.

According to another embodiment, the generating of the low-resolution difference images may include: determining an average value of pixel values in a pixel region for each patch region in the difference image; and generating a low-resolution difference image based on the average value corresponding to each patch region.

The pre-processed phase images include a pre-processed first phase image and a pre-processed second phase image, and the generating of the difference images may include shifting the pre-processed second phase image with different movement displacements. generating second phase images; and difference images representing a pixel value difference between each of the shifted second phase images and the preprocessed first phase image, and a difference representing a pixel value difference between the preprocessed second phase image and the preprocessed first phase image It may include generating an image.

The generating of the minimum map image may include: identifying a minimum difference value among difference values of regions corresponding to each other between the low-resolution difference images; and determining a pixel value of the minimum map image based on the minimum difference value.

A liveness test apparatus according to an embodiment includes a processor; and a memory including instructions executable by the processor, wherein when the instructions are executed by the processor, the processor performs preprocessing including edge enhancement processing on phase images of different phases to obtain the preprocessed phase generating images, generating difference images based on the preprocessed phase images, generating low-resolution difference images having a lower resolution than the difference images, and generating a minimum map image based on the low-resolution difference images, , a liveness check may be performed on the object displayed in the phase images based on the minimum map image.

According to an embodiment, an electronic device includes: a multi-phase detection sensor configured to acquire phase images of different phases; and a processor for performing a liveness check based on the phase images, wherein the processor performs preprocessing including edge enhancement processing on phase images of different phases to generate preprocessed phase images, Generate difference images based on the preprocessed phase images, generate low-resolution difference images having a lower resolution than the difference images, generate a minimum map image based on the low-resolution difference images, and Based on this, a liveness check may be performed on the object displayed in the phase images.

1 and 2 are diagrams for explaining biometric authentication and liveness check according to an embodiment.
3 is a view for explaining a multi-phase detection sensor according to an embodiment.
4 is a diagram for explaining a difference between a 2D object and a 3D object that can be detected through phase images according to an exemplary embodiment.
5 is a flowchart illustrating an operation of a liveness checking method according to an embodiment.
6 is a diagram for explaining a liveness test process according to an embodiment.
7 is a diagram for describing an edge enhancement process according to an embodiment.
8 and 9 are diagrams for explaining a process of generating a minimum map image according to an exemplary embodiment.
10A and 10B are diagrams for explaining a process of shifting a phase image according to an exemplary embodiment.
11A and 11B are diagrams for explaining a process of determining a pixel value of a minimum map image according to an exemplary embodiment.
12 is a diagram illustrating a configuration of a liveness testing apparatus according to an exemplary embodiment.
13 is a diagram illustrating a configuration of an electronic device according to an exemplary embodiment.

Specific structural or functional descriptions of the embodiments are disclosed for purposes of illustration only, and may be changed and implemented in various forms. Accordingly, the actual implemented form is not limited to the specific embodiments disclosed, and the scope of the present specification includes changes, equivalents, or substitutes included in the technical idea described in the embodiments.

Although terms such as first or second may be used to describe various elements, these terms should be interpreted only for the purpose of distinguishing one element from another. For example, a first component may be termed a second component, and similarly, a second component may also be termed a first component.

When a component is referred to as being “connected to” another component, it may be directly connected or connected to the other component, but it should be understood that another component may exist in between.

The singular expression includes the plural expression unless the context clearly dictates otherwise. In this specification, terms such as "comprise" or "have" are intended to designate that the described feature, number, step, operation, component, part, or combination thereof exists, and includes one or more other features or numbers, It should be understood that the possibility of the presence or addition of steps, operations, components, parts or combinations thereof is not precluded in advance.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related art, and should not be interpreted in an ideal or excessively formal meaning unless explicitly defined in the present specification. does not

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. In the description with reference to the accompanying drawings, the same components are assigned the same reference numerals regardless of the reference numerals, and overlapping descriptions thereof will be omitted.

1 and 2 are diagrams for explaining biometric authentication and liveness check according to an embodiment.

Biometrics is an authentication technology using personal biometric information such as a fingerprint, iris, face, vein, and skin among authentication technologies for user verification. In biometric authentication, face verification is an authentication method for determining whether a corresponding user is a valid user based on face information of a user who has attempted authentication. Facial authentication can be used to authenticate a valid user in user login, payment service, and access control.

Referring to FIG. 1 , an electronic device 120 (eg, the electronic device 1300 of FIG. 13 ) performs an authentication process for an object 110 (eg, a user) who wants to access the electronic device 120 through biometric authentication. carry out The electronic device 120 uses the image sensor 130 (eg, the 2PD sensor 310 of FIG. 3 or the multi-phase detection sensor 1240 of FIG. 12 ) included in the electronic device 120 . to obtain image data for the object 110, and analyze the obtained image data to determine an authentication result. The biometric authentication process may include extracting a feature from image data, comparing the extracted feature with a registered feature of a valid object, and determining whether authentication succeeds according to the comparison result. In an embodiment, when the electronic device 120 is in the locked state, when it is determined that authentication of the object 110 is successful, the locked state of the electronic device 120 may be released. Conversely, when it is determined that authentication of the object 110 has failed, the electronic device 120 continuously operates in the lock mode. The electronic device 120 may be a smartphone, a wearable device, a tablet computer, a netbook, a laptop, a desktop, a personal digital assistant (PDA), a set-top box, a home appliance, a biometric door lock, a security device, or a vehicle ignition device. .

A valid user of the electronic device 120 may register his/her biometric characteristics in the electronic device 120 in advance through a registration process, and the electronic device 120 stores the valid corresponding valid information in a storage device or cloud storage medium. Information for identifying the user may be stored. A face image of a valid user or a facial feature extracted from the face image may be stored as a registered biometric feature of the valid user.

In the biometric authentication process as described above, a liveness test may be performed. The liveness check is to check whether the object 110 is a living object, and is for determining whether the authentication means is authentic. The liveness test is to check whether the face shown in the image captured by the image sensor 130 is a genuine face or a fake face of a person. Liveness tests are used to distinguish between non-living objects (eg, photos, paper, video, models, and masks) and living objects (eg, real human faces). The electronic device 120 may perform either one of the liveness test and the biometric authentication, or both the liveness test and the biometric authentication, according to an embodiment.

2 shows a fake face 210 and a real face 220 according to one embodiment. The electronic device 120 may identify whether the face shown in the image data is the real face 220 through the liveness test. In addition, through the liveness test, the electronic device 120 detects a false face ( 210) can be distinguished.

An invalid user may use spoofing techniques to attempt to induce false acceptance of the biometric authentication system. In face authentication, an invalid user may present a photo or image showing a valid user's face to the image sensor 130 in order to cause mis-authentication. The liveness check prevents false authentication by filtering out authentication attempts (or spoofing attacks) using substitutes such as photos or videos. If the object to be authenticated is determined to be a non-living object as a result of the liveness check, it is finally determined that authentication has failed without performing the authentication process to determine whether it is a valid object, or even if the authentication process has been performed, regardless of the authentication result Finally, it may be determined that authentication has failed.

The image sensor 130 may express visual information on the object 110 as image data for a plurality of phases. The image sensor 130 may detect visual information of a plurality of phases and generate image data related to visual information of each phase. The image sensor 130 is a multi-phase detection sensor, and may be a two phase detection (2PD) sensor that detects two types of phases or a quad phase detection (QPD) sensor that detects four types of phases. However, the number of phases detected by the image sensor 130 is not limited thereto, and the image sensor 130 may detect other various numbers of phases. Hereinafter, embodiments in which the image sensor 130 corresponds to a 2PD sensor will be described for convenience of description, but the scope of the embodiment is not limited to the 2PD sensor, and the following description will be provided in which the image sensor 130 is a different type of sensor. It can also be applied to a case corresponding to a multi-phase detection sensor.

Referring to FIG. 3 , a structure of a 2PD sensor 310 is shown as an example of a multi-phase detection sensor. The 2PD sensor 310 includes a plurality of sensor pixels 315 capable of sensing color values of R (red), G (green), and B (blue) in two different phases. Reference numeral 320 is an enlarged representation of the sensor pixel 315 . Each sensor pixel 315 includes sensor sub-pixels 332 , 342 , 336 , 346 sensing a color value of G, sensor sub-pixels 334 , 344 sensing a color value of R and B sensor sub-pixels 338 , 348 for sensing color values. Here, the sensor sub-pixels 332 , 334 , 336 , 338 correspond to a first phase (eg, left (L)), and the sensor sub-pixels 342 , 344 , 346 , 348 correspond to the first phase. It may correspond to two phases (eg, right (R)). The structure of the 2PD sensor 310 illustrated in FIG. 3 is only an example, and the arrangement structure of the sensor pixels 315 is the example illustrated in FIG. 3 . It is not limited and may have various arrangement structures.

Each sensor sub-pixel included in the 2PD sensor 310 may sense visual information of any one of the first phase and the second phase. Each of the sensor sub-pixels 332 , 334 , 336 , 338 , 342 , 344 , 346 , and 348 includes a photodiode that receives external light and outputs an electrical signal value. When the output values of the photodiodes of the sensor sub-pixels 332 , 334 , 336 , 338 corresponding to the first phase are extracted, a first phase image 350 of the first phase is obtained, and the first phase image 350 of the first phase is obtained. When the output values of the photodiodes of the sensor sub-pixels 342 , 344 , 346 , and 348 are extracted, the second phase image 360 of the second phase may be obtained.

A disparity may exist between the first phase image 350 and the second phase image 360 , and accuracy of the liveness test may be improved by using the disparity for the liveness test of the object 110 . Since the parallax can represent three-dimensional information of the object 110, by using the parallax, it is possible to effectively distinguish a fake object implemented in 2D such as a photograph or paper. 4 is a diagram for explaining a difference between a 2D object and a 3D object that can be detected through phase images according to an exemplary embodiment. Referring to FIG. 4 , when a 2D object is photographed by the 2PD sensor 310 , a parallax is not detected through the first phase image and the second phase image. If the photo on which the face is printed is taken by the 2PD sensor 310, since there is no depth difference between the face region and the background region in the photo, and there is no depth difference between face parts, the first phase image and the second phase Parallax is not detected through the image. On the other hand, when the 3D object is photographed by the 2PD sensor 310 , a parallax may be detected through the first phase image and the second phase image. If a real human face is photographed by the 2PD sensor 310 , a disparity may be detected in the face contour and/or the nose region by comparing the first phase image and the second phase image.

Returning to FIG. 1 , through the phase images acquired through the image sensor 130 , the shape, color, texture, and context of the object 110 corresponding to 2D information are obtained. Information can be extracted. The electronic device 120 uses not only the phase images but also the disparity information derived from the phase images for the liveness test, and reflects the characteristics of the 3D structure of the object 110 in the liveness test result, thereby improving the accuracy of the liveness test. can improve The disparity information may be obtained from a differential image indicating a pixel value difference between phase images.

The amount of computation required for the liveness check and the accuracy of the liveness check are in conflict with each other to some extent. For high accuracy, information on the full resolution of possible phase images is required. However, assuming that the phase images have a resolution of megapixels, it is difficult to process the phase images of the full resolution in the mobile platform because processing the phase images of the full resolution is a very time-consuming and resource-consuming process. The electronic device 120 does not use the original phase images acquired by the image sensor 130 as it is, but rather in order to increase the processing speed of the liveness test and reduce resources (eg, memory) required for the liveness test. Liveness check is performed by converting images to low resolution. In the present specification, converting the phase image to a low resolution corresponds to reducing the size of the phase image. In particular, when the electronic device 120 is a mobile device such as a smart phone, it is important to reduce the weight of the liveness test, such as converting the resolution of the phase images to a low resolution, because there are large restrictions on the amount of computation and resources. However, the process of converting the phase images to a low resolution may cause loss of disparity-related information included in the phase images, thereby reducing the accuracy of disparity information and lowering the accuracy of the liveness test. In addition, the phase images acquired by the image sensor 130 may include a lot of noise caused by the movement of the electronic device 120 , and such noise prevents the acquisition of accurate disparity information, thereby reducing the accuracy of the liveness test. can be lowered

According to the liveness inspection method to be described below, by performing edge enhancement processing to enhance the edge component of the object 110 in phase images and effectively extracting main pixels from the difference image, accurate parallax even at low resolution make it possible to obtain information. Accuracy of the liveness test may be improved by accurate disparity information, and thus false authentication based on a forgery technique may be effectively prevented. In particular, when the liveness test is performed on a mobile platform such as a smartphone, the proposed liveness test method reduces the amount of computation and resources required in the liveness test process, enabling real-time processing to be sufficiently implemented. , the accuracy can be improved by performing liveness checks robustly against various spoofing attacks. Hereinafter, the liveness test method will be described in detail.

5 is a flowchart illustrating an operation of a liveness checking method according to an embodiment.

Referring to FIG. 5 , in operation 510 , the liveness testing apparatus (eg, the liveness testing apparatus 1200 of FIG. 12 ) receives phase images of different phases. The corresponding phase images may be obtained using a multi-phase detection sensor (eg, the multi-phase detection sensor 1240 of FIG. 12 ).

In operation 520, the liveness inspection apparatus generates preprocessed phase images by performing preprocessing including edge enhancement processing on the phase images. The liveness inspection apparatus removes noise from the phase images in a pre-processing process and performs a process of enhancing (or emphasizing) an edge region of an object appearing in each of the phase images. In an embodiment, the liveness checking apparatus may perform edge enhancement processing by applying a Sobel filter to each of the phase images. However, the edge enhancement processing is not limited to the method using the Sobel filter, and in addition, a method using one or more of an anisotropic filter, a Laplacian filter, and a canny edge filter, a neural network Various other edge enhancement processing, such as a method using , may be performed. In matching topological images, it is easier and more accurate to compare an edge region of an object (eg, the outline of a face) than a flat region (eg, a cheek of a face). Accordingly, information on the edge region is important in obtaining accurate disparity information. Through edge enhancement processing, such an edge region in phase images may be emphasized, and noise in a flat region may be removed from the shape of an object.

The liveness inspection apparatus may perform gamma correction on phase images before performing edge enhancement processing, and may perform edge enhancement processing on phase images on which gamma correction has been performed. Gamma correction is a process of correcting the brightness values of phase images using a nonlinear transfer function, and has an advantage that it can be performed with a low amount of calculation. The liveness inspection apparatus may also perform denoising processing for removing noise from phase images before performing edge enhancement processing.

In operation 530, the liveness test apparatus generates difference images based on the pre-processed phase images. The difference image is an image representing a difference in pixel values in pixel regions corresponding to each other between phase images. Assuming that the pre-processed phase images include the pre-processed first phase image and the pre-processed second phase image, the liveness test apparatus shifts the pre-processed second phase image by different shift displacements to be shifted. ) may generate second phase images. The liveness testing apparatus may generate difference images indicating a pixel value difference between each of the shifted second phase images and the preprocessed first phase image. The liveness test apparatus may calculate a pixel value difference in a corresponding region between the shifted second phase image and the preprocessed first phase image by shifting the preprocessed second phase image while the preprocessed first phase image is fixed. . Also, the liveness testing apparatus may generate a difference image indicating a pixel value difference between the non-shifted preprocessed second phase image and the preprocessed first phase image.

In operation 540, the liveness inspection apparatus generates low-resolution difference images having a lower resolution than the difference images. The liveness inspection apparatus may generate low-resolution difference images with reduced sizes by extracting main pixels from each of the difference images. The main pixels are pixels that are considered to contain important information or key signals among pixels of the difference image. For example, the liveness inspection apparatus may extract the largest pixel value for each patch region from the difference image according to a max pooling technique, and generate a low-resolution difference image based on the extracted largest pixel value. The largest pixel values extracted from the patch regions constitute pixel values of the low-resolution difference image. As another example, the liveness inspection apparatus determines an average value of pixel values in a pixel region for each patch region in a difference image according to an average pooling technique, and a low-resolution difference based on the average value corresponding to each patch region You can create an image. The average values determined in the patch regions constitute pixel values of the low-resolution difference image. However, the scope of the embodiment is not limited to the max pooling technique or the average pooling technique, and other techniques may be applied.

In operation 550, the liveness inspection apparatus generates a minimum map image based on the low-resolution difference images. The minimum map image is an image including a clue about 3D information of an object, similar to the disparity map. The liveness inspection apparatus may identify a minimum difference value among difference values of regions corresponding to each other between the low-resolution difference images. The liveness checking apparatus may determine a pixel value of the minimum map image based on the identified minimum difference value. A pixel value of a corresponding region in the minimum map image may correspond to a selected minimum difference value or may correspond to an index of a movement displacement corresponding to a low-resolution difference image including a minimum difference value among the difference images. The index of the displacement displacement indicates a displacement value shifted by the phase image. For example, the index of the translation displacement is +2 if shifted right by 2 pixels, +1 if shifted right 1 pixel, 0 if not shifted, -2 if shifted 2 pixels left, 1 pixel left It can have a value of -1 if it has been shifted by that much.

In operation 560, the liveness test apparatus performs a liveness test on the object shown in the phase images based on the minimum map image. The liveness test apparatus obtains a liveness test result by applying the minimum map image and the phase images to a neural network-based liveness test model (eg, the liveness test model 670 of FIG. 6 ). Here, the resolution of the phase images applied to the liveness test model may be adjusted to correspond to the resolution of the minimum map image. The minimum map image and phase images of the same resolution form input patches of the liveness inspection model. The minimum map image and the phase images may be concatenated to configure an input channel of the liveness test model.

The phase images included in the input patches may correspond to the entire region of the original phase images or may correspond to a partial region including a region of interest (eg, a face region).

The liveness check model may include one or more neural networks pretrained to detect the liveness of an object based on input patches. The neural network outputs values calculated by internal parameters in response to input data. At least a portion of the neural network may be implemented as software, hardware including a neural processor, or a combination of software and hardware. The neural network may correspond to a deep neural network (DNN) including a fully connected network, a deep convolutional network, and a recurrent neural network. have. A DNN may include a plurality of layers. The plurality of layers may include an input layer, one or more hidden layers, and an output layer. A neural network may be trained to perform a given operation by mapping input data and output data in a non-linear relationship to each other based on deep learning. Deep learning is a machine learning technique for solving a given problem from a big data set. Deep learning is a neural network optimization process that finds a point where energy is minimized while training a neural network using prepared training data.

One or more neural networks may output a liveness score in response to the input data. The liveness score is a standard value for determining whether an object is lively, and represents a numerical value, a probability value, or a feature value that the object corresponds to a real object or a false object. The liveness checking apparatus may determine whether the object is lively based on whether the liveness score satisfies a preset condition. When the liveness score is greater than the threshold value, the apparatus may determine that the object is a real living object, and when the liveness score is less than or equal to the threshold value, it may determine that the object is a non-living fake object.

As described above, the liveness test uses low-resolution phase images instead of using the original phase images obtained through the multi-phase detection sensor, thereby reducing the computational complexity and the amount of resources required for the liveness test, thereby enabling real-time processing. . In addition, the liveness check can improve the accuracy (or reliability) of the minimum map image by reducing the loss of 3D information in the process of generating a low-resolution minimum map image through the proposed edge enhancement processing and pooling technique. By effectively filtering out 2D spoofing attacks, the accuracy of the liveness check can be improved.

The liveness test apparatus may perform a control operation in response to a liveness test result of the object. When it is determined that the object is a real object, the liveness checking apparatus may request execution of a user authentication procedure. Conversely, when it is determined that the object is a fake object, the liveness checking apparatus may block the user's access without requesting execution of the user authentication process.

6 is a diagram for explaining a liveness test process according to an embodiment.

Referring to FIG. 6 , a first phase image 350 and a second phase image 360 are obtained from a multi-phase detection sensor such as the 2PD sensor 310 . In the liveness test apparatus (eg, the liveness test apparatus 1200 of FIG. 12 ), the first phase image 350 is first preprocessed by performing preprocessing including edge enhancement processing on the second phase image 360 , respectively. A phase image 610 and a preprocessed second phase image 620 are generated. The edge enhancement processing includes applying a filter that emphasizes an edge region of an object, such as a Sobel filter and an anisotropy filter, to the first phase image 350 and the second phase image 360 , respectively. According to an embodiment, the preprocessing may include gamma correction and/or denoising to remove noise.

The liveness testing apparatus generates difference images 630 based on the preprocessed first phase image 610 and the preprocessed second phase image 620 . The liveness inspection apparatus shifts the preprocessed second phase image 620 by different movement displacements to generate shifted second phase images, and generates shifted second phase images between the preprocessed first phase image 610 and the shifted second phase images. Difference images representing the pixel value difference may be generated. The difference images 630 may include a difference image indicating a pixel value difference between the unshifted preprocessed second phase image 620 and the preprocessed first phase image 610 .

The liveness inspection apparatus generates the low-resolution difference images 640 by reducing the size of the difference images 630 . The liveness inspection apparatus may generate the low-resolution difference images 640 by extracting main pixels from each of the difference images 630 using a max pooling technique.

The liveness inspection apparatus generates a minimum map image 650 based on the low-resolution difference images 640 . The liveness checking apparatus may identify a minimum difference value among difference values of corresponding regions between the low-resolution difference images 640 and determine a pixel value of the minimum map image 650 based on the minimum difference value. The liveness checking apparatus may determine the minimum difference value as the pixel value of the minimum map image 650 as it is, or determine the shift index of the low-resolution difference image having the minimum difference value as the pixel value of the minimum map image 650 .

The liveness inspection apparatus generates a low-resolution first phase image 660 and a low-resolution second phase image 665 by reducing the resolution of the first phase image 350 and the second phase image 360 , and the low-resolution first phase image The liveness test may be performed by applying the 660 , the low-resolution second phase image 665 , and the minimum map image 650 to the neural network-based liveness test model 670 . Instead of analyzing the full-resolution first phase image 350 and the second phase image 360 obtained from the 2PD sensor 310, the low-resolution first phase image 660 and the low-resolution second phase image 665 and It is possible to use a lightweight neural network as the liveness test model 670 because it is enough to analyze the minimum map image 650 with reduced resolution.

The low-resolution first phase image 660 , the low-resolution second phase image 665 , and the minimum map image 650 input to the liveness test model 670 may have the same resolution. The low-resolution first phase image 660 , the low-resolution second phase image 665 , and the minimum map image 650 constitute an input channel of the liveness test model 670 and pass through the input layer of the liveness test model 670 . may be input to the liveness test model 670 . According to an embodiment, the image size of the minimum map image 650 may be reduced to a resolution defined before being input to the liveness test model 670. In this case, the first phase image 350 and the second phase image Before 360 degrees are input to the liveness test model 670 , the image size may be reduced to be the same as the resolution of the minimum map image 650 .

The liveness test model 670 may output a liveness score in response to input data. The liveness checking apparatus may determine that the object is a real object if the liveness score satisfies a predefined condition (eg, a condition greater than a threshold value), and may determine that the object is a fake object if the condition is not satisfied.

7 is a diagram for describing an edge enhancement process according to an embodiment.

Referring to FIG. 7 , the liveness inspection apparatus (eg, the liveness inspection apparatus 1200 of FIG. 12 ) performs edge enhancement on each of the first phase image 350 and the second phase image 360 . The liveness inspection apparatus may apply the Sobel filter 710 to the first phase image 350 to perform preprocessing for emphasizing the edge region of the object displayed in the first phase image 350 . Also, the liveness testing apparatus may apply the Sobel filter 710 to the second phase image 360 to perform preprocessing for emphasizing the edge region of the object displayed in the second phase image 360 . Through such a pre-processing process, a pre-processed first phase image 610 and a pre-processed second phase image 620 in which noise is reduced and an edge region is emphasized may be obtained. According to an embodiment, an edge enhancement filter of an anisotropic filter, a Laplacian filter, or a Canny edge filter as well as the Sobel filter 710 may be used for edge enhancement processing. Through edge enhancement processing, it is possible to detect parallax information with higher accuracy in the edge area of an object.

8 is a diagram for explaining a schematic generation process of a minimum map image according to an embodiment.

Referring to FIG. 8 , after preprocessing including edge enhancement processing is performed, difference images 630 are generated based on the preprocessed first phase image 610 and the preprocessed second phase image 620 . . The liveness test apparatus (eg, the liveness test apparatus 1200 of FIG. 12 ) shifts the preprocessed second phase image 620 by different movement displacements to obtain the shifted second phase images and the preprocessed first phase. Difference images 630 representing the pixel value difference between the images 610 may be generated. Here, the difference images 630 include a first difference image 812 , a second difference image 814 , a third difference image 816 , a fourth difference image 818 , and a fifth difference image 820 . assume that The first difference image 812 represents a pixel value difference between the preprocessed first phase image 610 and the image obtained by shifting the preprocessed second phase image 620 to the left by 2 pixels, and the second difference image 814 . denotes a pixel value difference between the preprocessed first phase image 610 and the preprocessed second phase image 620 shifted by one pixel to the left. The third difference image 816 represents a pixel value difference between the preprocessed first phase 610 and the non-shifted preprocessed second phase image 620 . The fourth difference image 818 represents a pixel value difference between the preprocessed first phase image 610 and the preprocessed second phase image 620 shifted by one pixel to the right, and the fifth difference image 820 . denotes a pixel value difference between the preprocessed first phase image 610 and the preprocessed second phase image 620 shifted by 2 pixels to the right. However, in relation to generating the difference images 630 , the range is not limited to shifting the preprocessed second phase image 620 in the embodiment. In another embodiment, shifted first phase images are generated by shifting the preprocessed first phase image 610 with different movement displacements, and the preprocessed second phase image 620 and the shifted second phase images are respectively It is also possible to generate difference images representing the difference in pixel values between

After generating the difference images 630 , the liveness inspection apparatus generates low-resolution difference images 640 having a smaller resolution than the difference images 630 . In an embodiment, the liveness testing apparatus includes a first difference image 812 , a second difference image 814 , a third difference image 816 , a fourth difference image 818 , and a fifth difference image 820 , respectively. By applying the max pooling technique, the first low-resolution difference image 832, the second low-resolution difference image 834, the third low-resolution difference image 836, the fourth low-resolution difference image 838, and the fifth low-resolution difference image 840 ) can be created. The liveness inspection apparatus may generate the minimum map image 650 based on the smallest difference value among the pixel values in a region corresponding to each other between the low-resolution difference images 640 . The minimum map image 650 may have the smallest difference value as a pixel value or an index of movement displacement of the low-resolution difference image including the smallest difference value as a pixel value.

9 is a diagram for describing in more detail a process of generating a minimum map image according to an exemplary embodiment.

Referring to FIG. 9 , in operation 910 , the liveness checking apparatus performs a shift of the phase image. The liveness inspection apparatus may shift the XNth phase image while the first phase image is fixed. In FIG. 9 , a number in each pixel of each phase image indicates a pixel value. In the 'XNth phase image', 'X' indicates that phase characteristics are divided in the horizontal direction, and 'N' indicates the number of phases. For example, when the first phase image and the second phase image generated by the 2PD sensor are used, the second phase image may be displayed as an X2 th phase image. Hereinafter, an embodiment in which the XNth phase image corresponds to the second phase image will be described. The liveness testing apparatus may set a basic area in the first phase image and set one or more shift areas in the second phase image. For example, the liveness checking apparatus may set shift regions of x-1, x0, and x+1 in the second phase image. Here, x0 represents a basic area in which no shift is performed.

The basic region of the first phase image may be referred to as a first basic region, the basic region of the second phase image may be referred to as a second basic region, and the first basic region and the second basic region may be positioned with each other in position. can respond In x-1 and x+1, - and + indicate a shift direction, respectively, and 1 indicates a reference shift value. The basic area may be set based on a reference shift value. When the reference shift value is r, a shift region in which the basic region is shifted by r in a specific direction is set.

The liveness checking apparatus shifts the second reference region (ie, the shift region of x0) by the reference shift value (ie, 1) along the shift direction to one or more shift regions (ie, the shift region of x-1, and x+ 1) can be set. The reference shift value may be set to various values, and the number of shift regions corresponding to the reference shift value may be set. The number of shift regions may be determined based on the reference shift value and the number of shift directions.

For example, when the reference shift value is 1 and the number of shift directions is 2 (left and right), 2×1+1=3 shift regions may exist. The three shift regions may include shift regions of x-1, x0, and x+1. As another example, when the reference shift value is 5 and the number of shift directions is 2 (left and right), 2×5+1=11 shift regions may exist. The 11 shift regions may include shift regions of x-5 to x-1, x0, and x+1 to x+5. As another example, when the reference shift value is 1 and the number of shift directions is four (left, right, upper, lower), 2×1+2×1+1=5 shift regions may exist. The five shift regions may include shift regions of x-1, y-1, xy0, x+1, and y+1.

When a multi-phase detection sensor such as a QPD sensor is used, phase characteristics may be distinguished in directions other than the horizontal direction. The liveness test apparatus may determine shift regions for each phase image by performing shifting on the phase images in the horizontal and vertical directions of the QPD sensor as shown in FIG. 10A . In the case of the XNth phase image, shift regions x-1, x0, and x+1 may be determined through a horizontal shift like the phase image shift 910 of FIG. 9 . In the case of the YN-th phase image, shift regions y-1, y0, and y+1 may be determined through vertical shift.

In XN and YN, 'X' indicates that phase characteristics are divided in a horizontal direction, and 'Y' indicates that phase characteristics are divided in a vertical direction. 'N' represents the number of phases. Although it is described herein that the same number of phases are used in the vertical and horizontal directions, it is also possible that different numbers of phases are used in the vertical and horizontal directions. 'N' may be determined based on the number of phases that the sensor can distinguish. In the case of the QPD sensor, N=2, and in the embodiment of FIG. 10A , a first phase image, an X2th phase image, and a Y2th phase image may exist.

According to another embodiment, the liveness testing apparatus may determine shift regions for each phase image by performing a phase image shift on the phase images in the horizontal direction, the vertical direction, and the diagonal direction of the QPD sensor as shown in FIG. 10B . . In the case of an XN-th phase image, shift regions (x-1, x0, and x+1) may be determined through a shift in the horizontal direction, and in the case of a YN-th phase image, shift regions y-1 through a shift in the vertical direction. , y0 and y+1) can be determined. In the case of the ZN-th phase image, shift regions z-1, z0, and z+1 may be determined through diagonal shift. In 'ZN', 'Z' indicates that phase characteristics are divided in a diagonal direction, and 'N' indicates the number of phases. When N=2, the first phase image, the X2th phase image, the Y2th phase image, and the Z2th phase image may be used in the embodiment of FIG. 10B .

When the shift regions are determined in this way, in operation 920 , the liveness checking apparatus calculates differences between the image of the basic region and the image of each shift region. The liveness inspection apparatus generates difference images based on a difference between a fixed image (eg, an image of the first basic region) and a shifted image (eg, an image of a shift region), and a minimum map image based on the difference images can create The liveness checking apparatus generates a first difference image based on a difference between the image of the first basic region and the image of the shift region of x-1, and based on the difference between the image of the first basic region and the image of the shift region of x0 Thus, a second difference image may be generated, and a third difference image may be generated based on a difference between the image of the first basic region and the image of the x+1 shift region. The liveness test apparatus may assign an index value to each difference image. For example, the liveness checking apparatus may assign index values in the order of x-1, x0, and x+1. In FIG. 9 , it is assumed that an index value of 0 is assigned to the primary image, an index value of 1 is assigned to the secondary image, and an index value of 2 is assigned to the tertiary image. However, the scope of the embodiment is not limited thereto, and indexes may be assigned in other various orders.

The liveness test apparatus uses a pooling technique that selects only some pixels from the first low-resolution difference image, the second low-resolution difference image, and the third low-resolution image from each of the first, second, and tertiary images. You can create a difference image.

In operation 930, the liveness testing apparatus generates a minimum map image. The liveness checking apparatus may identify a minimum difference value among corresponding difference values of regions corresponding to each other in the low-resolution difference images, and determine a pixel value of the minimum map image based on the minimum difference value. For example, in the embodiment of FIG. 9 , corresponding difference values located at (1, 1) are 1, 0, and 6. Among them, 0 may be selected as the minimum difference value. As another example, the corresponding difference values located at (2, 2) are 25, 33, 30. Among them, 25 may be selected as the minimum difference value. As such, a minimum difference value may be selected among corresponding difference values between the low-resolution difference images, and a pixel value of the minimum map image may be determined based on the minimum difference value.

The pixel value of the minimum map image may correspond to a minimum difference value or may correspond to an index of a low-resolution difference image including a minimum difference value among low-resolution difference images. The minimum map image including the minimum difference value may be referred to as a minimum difference value map image 934 , and the minimum map image including the index of the low-resolution difference image including the minimum difference value is the minimum index map image 932 . may be referred to. In the previous example, 0 is selected as the minimum difference value at the position of (1, 1), and the index of the low-resolution difference image including 0 is 1. Accordingly, the pixel value of (1, 1) in the minimum difference value map image 934 is 0, and the pixel value of (1, 1) in the minimum index map image 932 is 1. In addition, at the position of (2, 2), 25 was selected as the minimum difference value, and the index of the low-resolution difference image including 25 is 0. Accordingly, the pixel value of (2, 2) in the minimum difference value map image 934 is 25, and the pixel value of (2, 2) in the minimum index map image 932 is 0.

A set of difference images may be generated for each phase image. As in the embodiments of FIGS. 10A and 10B , when phase images related to a plurality of directions exist, a minimum map image of each phase image may be generated based on a set of difference images of each phase image. In the embodiment of FIG. 10A , a minimum map image for each of the XNth phase image and the YNth phase image may be generated. 10B , a minimum map image for each of the XNth phase image, the YNth phase image, and the ZNth phase image may be generated.

11A and 11B are diagrams for explaining a process of determining a pixel value of a minimum map image according to an exemplary embodiment.

Referring to FIG. 11A , a region 1132 of a first low-resolution difference image 832 , a region 1134 of a second low-resolution difference image 834 , a region 1136 of a third low-resolution difference image 836 , and a fourth The region 1138 of the low-resolution difference image 838 and the region 1140 of the fifth low-resolution difference image 840 correspond to each other. The liveness testing apparatus identifies the smallest difference value among the difference values represented by the regions 1132 , 1134 , 1136 , 1138 , and 1140 . If it is assumed that the difference value of the region 1134 is smaller than the difference values of the other regions 1132 , 1136 , 1138 , and 1140 , the liveness test apparatus calculates the difference value of the region 1134 from the minimum map image 650 . It may be determined as a pixel value of the region 1150 corresponding to the region 1134 . Alternatively, the liveness checking apparatus determines a shift index (eg, -1) corresponding to the third low-resolution difference image 836 including the area 1134 as a pixel value of the area 1150 of the minimum map image 650 . may be

Referring to FIG. 11B , an area 1162 of the first low-resolution difference image 832 , an area 1164 of the second low-resolution difference image 834 , an area 1166 of the third low-resolution difference image 836 , and a fourth The region 1168 of the low-resolution difference image 838 and the region 1170 of the fifth low-resolution difference image 840 correspond to each other. The liveness testing apparatus identifies the smallest difference value among the difference values indicated by the regions 1162 , 1164 , 1166 , 1168 , and 1170 similarly to FIG. 11A . If it is assumed that the difference value of the region 1168 is smaller than the difference values of the other regions 1162 , 1164 , 1166 and 1170 , the liveness test apparatus calculates the difference value of the region 1168 from the minimum map image 650 . It may be determined as a pixel value of the region 1180 corresponding to the region 1168 . Alternatively, the liveness checking apparatus determines a corresponding shift index (eg, +1) of the fourth low-resolution difference image 838 including the area 1168 as a pixel value of the area 1180 of the minimum map image 650 . may be

The liveness inspection apparatus may determine the minimum map image 650 by performing the above process for all corresponding regions of the low-resolution difference images 640 . When pixel values of the minimum map image 650 are configured using the minimum difference values, a minimum difference value map image (eg, the minimum difference value map image 934 of FIG. 9 ) is generated, and the minimum map image 650 is generated using the shift indices. By configuring the pixel values of , a minimum index map image (eg, the minimum index map image 932 of FIG. 9 ) is generated. After the minimum map image 650 is determined, the minimum map image 650 and the phase images adjusted to the same resolution as the resolution of the minimum map image 650 may be input to the liveness test model.

12 is a diagram illustrating a configuration of a liveness testing apparatus according to an exemplary embodiment.

Referring to FIG. 12 , the liveness test apparatus 1200 includes a processor 1210 and a memory 1220 . The memory 1220 is connected to the processor 1210 and may store instructions executable by the processor 1210 , data to be operated by the processor 1210 , or data processed by the processor 1210 . Memory 1220 may include high-speed random access memory and/or non-volatile computer-readable storage media (eg, flash memory).

The processor 1210 may execute instructions for executing one or more operations described with reference to FIGS. 1 to 11B . The processor 1210 may receive phase images of different phases obtained by the multi-phase detection sensor 1240 , and perform preprocessing including edge enhancement processing on the phase images to generate preprocessed phase images. The processor 1210 may perform a process of removing noise from the phase images and reinforcing an edge region of an object displayed in each of the phase images. The processor 1210 may generate difference images based on the pre-processed phase images, and may generate low-resolution difference images having a lower resolution than the difference images. The processor 1210 may extract the largest pixel value for each patch region from the difference image, and generate a low-resolution difference image based on the extracted largest pixel value. As another example, the processor 1210 may determine an average value of pixel values in a pixel region for each patch region in the difference image, and generate a low-resolution difference image based on the average value corresponding to each patch region.

The processor 1210 generates a minimum map image based on the low-resolution difference images. The processor 1210 may identify a minimum difference value among difference values of regions corresponding to each other between the low-resolution difference images, and determine a pixel value of the minimum map image based on the identified minimum difference value. The pixel value of the minimum map image may correspond to a selected minimum difference value or may correspond to an index of a movement displacement corresponding to a low-resolution difference image including a minimum difference value among the difference images. The processor 1210 may perform a liveness check on the object displayed in the phase images based on the minimum map image. The processor 1210 may obtain a liveness test result by applying the minimum map image and the phase images to a neural network-based liveness test model (eg, the liveness test model 670 of FIG. 6 ). The resolution of the phase images applied to the liveness test model may be adjusted to correspond to the resolution of the minimum map image. The liveness test model may output a liveness score as a test result, and the processor 1210 may determine whether the object is a real object or a fake object based on a result of comparison between the liveness score and the threshold value.

13 is a diagram illustrating a configuration of an electronic device according to an exemplary embodiment.

Referring to FIG. 13 , the electronic device 1300 (eg, the electronic device 120 of FIG. 1 ) generates an input image (eg, a phase image) obtained by photographing an object, and determines whether the object displayed in the input image is live. can be inspected Also, the electronic device 1300 may perform biometric authentication (eg, face authentication and iris authentication) based on the liveness test result of the object. The electronic device 1300 may perform a function of the liveness test apparatus 1200 of FIG. 12 .

The electronic device 1300 may include a processor 1310 , a memory 1320 , an image sensor 1330 , a storage device 1350 , an input device 1360 , an output device 1370 , and a communication device 1380 . . The processor 1310 , the memory 1320 , the image sensor 1330 , the storage device 1350 , the input device 1360 , the output device 1370 , and the communication device 1380 communicate with each other through the communication bus 1390 . can

The processor 1310 executes functions and instructions for execution in the electronic device 1300 . The processor 1310 may process instructions stored in the memory 1320 or the storage device 1350 . The processor 1310 may perform one or more operations described with reference to FIGS. 1 to 12 . The processor 1310 generates preprocessed phase images by performing preprocessing including edge enhancement processing on phase images of different phases obtained through the image sensor 1330 , and a difference image based on the preprocessed phase images can create The processor 1310 may generate low-resolution difference images having a lower resolution than the difference images, and generate a minimum map image based on the low-resolution difference images. The processor 1310 performs a liveness check on the object shown in the phase images based on the minimum map image. The processor 1310 obtains a liveness score by applying the minimum map image and the phase images to a neural network-based liveness test model (eg, the liveness test model 670 of FIG. 6 ), and obtains a liveness score and a threshold value. A liveness test result may be determined based on a comparison result between livers.

The memory 1320 stores data for performing a liveness test. The memory 1320 may store instructions for execution by the processor 1310 and information for performing a liveness check and/or biometric authentication.

The image sensor 1330 may generate an image by photographing an object. The image sensor 1330 may include a multi-phase detection sensor (eg, a 2PD sensor or a QPD sensor) that acquires phase images of different phases.

The storage device 1350 includes a computer-readable storage medium. The storage device 1350 may store a larger amount of information than the memory 1320 and may store the information for a long period of time. The storage device 1350 may include a magnetic hard disk, an optical disk, a flash memory, or a floppy disk.

The input device 1360 may receive an input from the user through tactile, video, audio, or touch input. The input device 1360 may include a keyboard, mouse, touch screen, microphone, or any other device capable of detecting input from a user and passing the detected input to the electronic device 1300 .

The output device 1370 may provide an output of the electronic device 1300 to the user through a visual, auditory, or tactile channel. Output device 1370 may include a display, touch screen, speaker, vibration generating device, or any other device capable of providing output to a user. The communication device 1380 may communicate with an external device through a wired or wireless network.

The embodiments described above may be implemented by a hardware component, a software component, and/or a combination of a hardware component and a software component. The apparatus, method and components described in the embodiments include a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), a programmable logic unit (PLU) ), a microprocessor, or any other device capable of executing and responding to instructions, may be implemented using a general purpose computer or special purpose computer. The processing device may execute an operating system (OS) and a software application executed on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For convenience of understanding, although one processing device is sometimes described as being used, one of ordinary skill in the art will recognize that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that can include The processing device may include a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as parallel processors.

Software may comprise a computer program, code, instructions, or a combination of one or more thereof, which configures a processing device to operate as desired or is independently or collectively processed You can command the device. The software and/or data may be any kind of machine, component, physical device, virtual equipment, computer storage medium or apparatus, to be interpreted by or to provide instructions or data to the processing device. , or may be permanently or temporarily embody in a transmitted signal wave. The software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored in a computer-readable recording medium.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination, and the program instructions recorded on the medium are specially designed and configured for the embodiment, or are known and available to those skilled in the art of computer software. may be Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic such as floppy disks. - includes magneto-optical media, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like.

The hardware devices described above may be configured to operate as one or a plurality of software modules to perform the operations of the embodiments, and vice versa.

As described above, although the embodiments have been described with reference to the limited drawings, those of ordinary skill in the art may apply various technical modifications and variations based thereon. The described techniques are performed in an order different from the described method, and/or the described components of the system, structure, apparatus, circuit, etc. are combined or combined in a different form than the described method, or other components or equivalents Appropriate results can be achieved even if substituted or substituted by

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (20)

In the liveness test method,
receiving phase images of different phases;
generating preprocessed phase images by performing preprocessing including edge enhancement processing on the phase images;
generating differential images based on the preprocessed phase images;
generating low-resolution difference images having a lower resolution than the difference images;
generating a minimum map image based on the low-resolution difference images; and
performing a liveness check on the object displayed in the phase images based on the minimum map image
A liveness test method comprising a.
According to claim 1,
Performing the pre-processing step,
removing noise from the phase images and performing a process of strengthening an edge region of an object displayed in each of the phase images
A liveness test method comprising a.
According to claim 1,
Performing the pre-processing step,
performing edge enhancement processing by applying at least one of a Sobel filter, an anisotropic filter, a Laplacian filter, and a Canny edge filter to each of the phase images
A liveness test method comprising a.
According to claim 1,
The step of generating the low-resolution difference images comprises:
extracting the largest pixel value for each patch region from the difference image; and
generating a low-resolution difference image based on the extracted largest pixel value
A liveness test method comprising a.
According to claim 1,
The step of generating the low-resolution difference images comprises:
determining an average value of pixel values within a pixel area for each patch area in the difference image; and
generating a low-resolution difference image based on the average value corresponding to each patch region;
A liveness test method comprising a.
According to claim 1,
The pre-processed phase images are
comprising a preprocessed first phase image and a preprocessed second phase image,
The step of generating the difference images comprises:
generating shifted second phase images by shifting the preprocessed second phase image by different shift displacements; and
Difference images representing a pixel value difference between each of the shifted second phase images and the preprocessed first phase image, and a difference image representing a pixel value difference between the preprocessed second phase image and the preprocessed first phase image steps to create
A liveness test method comprising a.
According to claim 1,
The step of generating the minimum map image comprises:
identifying a minimum difference value among difference values of regions corresponding to each other between the low-resolution difference images; and
determining a pixel value of the minimum map image based on the minimum difference value;
A liveness test method comprising a.
8. The method of claim 7,
The pixel value of the minimum map image is,
Corresponding to the index of movement displacement corresponding to the minimum difference value or corresponding to the low-resolution difference image including the minimum difference value among the low-resolution difference images,
How to check liveness.
According to claim 1,
The step of performing the liveness test comprises:
obtaining a liveness test result by applying the minimum map image and the phase images to a neural network-based liveness test model;
A liveness test method comprising a.
10. The method of claim 9,
The resolution of the phase images applied to the liveness test model is adjusted to correspond to the resolution of the minimum map image,
How to check liveness.
According to claim 1,
The phase images of the different phases are obtained using a multi-phase detection sensor,
How to check liveness.
A computer-readable recording medium storing one or more computer programs including instructions for performing the method of any one of claims 1 to 11.
A liveness test apparatus comprising:
processor; and
memory containing instructions executable by the processor
including,
When the instructions are executed by the processor, the processor:
Performing preprocessing including edge enhancement processing on phase images of different phases to generate preprocessed phase images,
generating difference images based on the pre-processed phase images;
generating low-resolution difference images having a lower resolution than the difference images;
Generate a minimum map image based on the low-resolution difference images,
performing a liveness check on the object shown in the phase images based on the minimum map image,
Liveness check device.
14. The method of claim 13,
The processor is
removing noise from the phase images and performing a process of strengthening an edge region of an object appearing in each of the phase images,
Liveness check device.
14. The method of claim 13,
The processor is
extracting the largest pixel value for each patch region from the difference image, and generating a low-resolution difference image based on the extracted largest pixel value,
Liveness check device.
14. The method of claim 13,
The processor is
determining an average value of pixel values in a pixel region for each patch region in the difference image, and generating a low-resolution difference image based on the average value corresponding to each patch region,
Liveness check device.
14. The method of claim 13,
The processor is
identifying a minimum difference value among difference values of regions corresponding to each other between the low-resolution difference images, and determining a pixel value of the minimum map image based on the minimum difference value;
Liveness check device.
14. The method of claim 13,
The processor is
obtaining a liveness test result by applying the minimum map image and the phase images to a neural network-based liveness test model;
The resolution of the phase images applied to the liveness test model is adjusted to correspond to the resolution of the minimum map image,
Liveness check device.
An electronic device for performing a liveness check of an object, the electronic device comprising:
a multi-phase detection sensor that acquires phase images of different phases; and
A processor that performs a liveness check based on the phase images
including,
The processor is
Performing preprocessing including edge enhancement processing on phase images of different phases to generate preprocessed phase images,
generating difference images based on the pre-processed phase images;
generating low-resolution difference images having a lower resolution than the difference images,
Generate a minimum map image based on the low-resolution difference images,
performing a liveness check on the object shown in the phase images based on the minimum map image,
electronic device.
20. The method of claim 19,
The processor is
removing noise from the phase images and performing a process of strengthening an edge region of an object displayed in each of the phase images;
extracting the largest pixel value for each patch region from the difference image, and generating a low-resolution difference image based on the extracted largest pixel value,
electronic device.

Images