Face recognition method, device and equipment based on euler image magnification and medium

Abstract

The present disclosure provides a face recognition method based on Euler image magnification, which can be used in the field of artificial intelligence technology. The method comprises: acquiring video data containing a face and light control information, wherein the light control information comprises a time sequence associated with the video data; determining a cheek region and a pupil region in face data according to the video data; magnifying the change information of the face data through an Euler image magnification algorithm to determine first change information associated with the cheek region and second change information associated with the pupil region; and determining a face recognition result according to the first change information, the second change information and the light control information. The present disclosure also provides a face recognition device based on Euler image magnification, an electronic device and a storage medium.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, and more specifically, to a face recognition method, apparatus, electronic device, computer-readable storage medium, and computer program product based on Euler image magnification. Background Technology

[0002] In various security scenarios such as password modification and identity authentication in banking, it is necessary to authenticate the identity of relevant users, such as by recording facial information. This is typically done by setting up designated service locations and using manual data collection and verification. For example, if a user needs to authenticate their identity or handle authorization-related matters, they go to a designated service location and are guided by staff to record their facial information. Alternatively, an online and manual collaborative model can be used: the application is submitted online, facial data is collected offline, the application is submitted for review online, and the progress is checked online. The identity verification is primarily handled by branch staff. Another option is an interactive online model, where the customer completes the entire online application process, requiring interactive verification of facial information and other identity details.

[0003] However, in existing technologies, users need to complete interactive tasks according to preset actions during the face registration process, which requires users to perform multiple verifications. In addition, this type of verification method has low security and is easily cracked through related video methods, making it impossible to effectively determine the authenticity of the face. Summary of the Invention

[0004] In view of the above problems, this disclosure provides a face recognition method, device, electronic device, readable storage medium, and computer program product based on Euler image magnification. It can automatically determine the authenticity of a face based on the user's verification video data without requiring the user to perform a verification operation according to the set actions. In addition, during the face recognition process, the face recognition result can be determined by first change information related to the cheek, second change information related to the pupil, and light control information in the face data, which can effectively improve the accuracy and security of face recognition.

[0005] One aspect of this disclosure provides a face recognition method based on Euler image magnification, comprising: acquiring video data containing a face and lighting control information, the lighting control information including a time series associated with the video data; determining a cheek region and a pupil region in the face data based on the video data; magnifying the change information of the face data using an Euler image magnification algorithm to determine first change information associated with the cheek region and second change information associated with the pupil region; and determining a face recognition result based on the first change information, the second change information, and the lighting control information.

[0006] In some embodiments of this disclosure, determining a face recognition result based on the first change information, the second change information, and the light control information includes: determining whether the first change information meets a first set threshold and whether the second change information meets a second set threshold using a pre-trained judgment model; comparing whether the second change information and the light control information have a change consistency; if the first change information meets the first set threshold, the second change information meets the second set threshold, and the second change information and the light control information have a change consistency, then the face recognition result is confirmed to be true; otherwise, the face recognition result is determined to be false.

[0007] In some embodiments of this disclosure, the Euler image magnification algorithm is used to amplify the change information of the face data, determine the first change information associated with the cheek region and the second change information associated with the pupil region, including: performing spatial filtering and downsampling on the cheek region and pupil region in the face data to generate a first baseband and a second baseband; performing bandpass filtering on the first baseband and the second baseband in the time domain to extract the first change sub-information and the second change sub-information; amplifying the first change sub-information and the second change sub-information to a set factor, and merging them with the first baseband and the second baseband respectively to determine the first change information and the second change information.

[0008] In some embodiments of this disclosure, spatial filtering and downsampling are performed on the cheek region and pupil region in the face data to generate a first baseband and a second baseband, including: obtaining a first pixel value matrix of the color channels in the cheek region and a second pixel matrix of the color channels in the pupil region; using a low-pass filter to spatially filter the first pixel value matrix to obtain first data with different spatial frequencies, and spatially filtering the second pixel matrix to obtain second data with different spatial frequencies; and smoothing and downsampling the first data and the second data according to a Gaussian pyramid to obtain the first baseband and the second baseband respectively.

[0009] In some embodiments of this disclosure, determining the cheek region and pupil region in face data based on the video data includes: generating face data based on the video data; and determining the cheek region and pupil region based on the face data.

[0010] In some embodiments of this disclosure, generating face data based on the video data includes: performing face segmentation on the video data to extract face regions in each video frame; and aligning the extracted face regions to generate face data.

[0011] In some embodiments of this disclosure, aligning the extracted face regions to generate face data includes: determining facial key points based on the face regions; performing geometric transformations on the face regions to place the facial key points in preset fixed positions; and aligning the geometrically transformed face regions in each video frame to generate face data.

[0012] In some embodiments of this disclosure, the facial key points include: eyebrow key points, eye key points, nose key points, mouth key points, and facial contour key points.

[0013] In some embodiments of this disclosure, determining the cheek region and the pupil region based on the facial data includes: determining the cheek region based on the eyebrow key points, eye key points, nose key points, mouth key points, and facial contour key points; and determining the pupil region based on the eye key points.

[0014] In some embodiments of this disclosure, the first set threshold includes the cycle of real face color change, and the second set threshold includes the cycle of real pupillary nerve reflex change.

[0015] Another aspect of this disclosure provides a face recognition device based on Euler image magnification, comprising: an acquisition module configured to acquire video data containing a face and light control information, wherein the light control information includes a time series associated with the video data; a first determination module configured to determine a cheek region and a pupil region in the face data based on the video data; a second determination module configured to magnify the change information of the face data using an Euler image magnification algorithm to determine first change information associated with the cheek region and second change information associated with the pupil region; and a third determination module configured to determine a face recognition result based on the first change information, the second change information, and the light control information.

[0016] In some embodiments of this disclosure, the third determining module includes a result generating module, configured to: determine whether the first change information meets a first set threshold and whether the second change information meets a second set threshold using a pre-trained judgment model; compare whether the second change information and the light control information have a change consistency; if the first change information meets the first set threshold, the second change information meets the second set threshold, and the second change information and the light control information have a change consistency, then confirm that the face recognition result is true; otherwise, determine that the face recognition result is false.

[0017] In some embodiments of this disclosure, the second determining module includes a change information determining module configured to: perform spatial filtering and downsampling on the cheek region and pupil region in the face data to generate a first baseband and a second baseband; perform bandpass filtering on the first baseband and the second baseband in the time domain to extract first change sub-information and second change sub-information; amplify the first change sub-information and the second change sub-information to a set factor, and merge them with the first baseband and the second baseband respectively to determine the first change information and the second change information.

[0018] In some embodiments of this disclosure, the change information determination module includes a baseband generation module configured to: acquire a first pixel value matrix of the color channel in the cheek region and a second pixel matrix of the color channel in the pupil region; perform spatial filtering on the first pixel value matrix using a low-pass filter to obtain first data with different spatial frequencies, and perform spatial filtering on the second pixel matrix to obtain second data with different spatial frequencies; and perform smoothing and downsampling on the first data and the second data according to a Gaussian pyramid to obtain the first baseband and the second baseband respectively.

[0019] In some embodiments of this disclosure, the first determining module includes a first determining unit configured to: generate face data based on the video data; and determine the cheek region and pupil region based on the face data.

[0020] In some embodiments of this disclosure, the first determining unit includes a face data generation module, configured to: perform face segmentation on the video data, extract face regions from each video frame; and align the extracted face regions to generate face data.

[0021] In some embodiments of this disclosure, the face data generation module includes a face data generation unit configured to: determine facial key points based on the face region; perform geometric transformation on the face region to place the facial key points in a preset fixed position; and align the geometrically transformed face regions in each video frame to generate face data.

[0022] In some embodiments of this disclosure, the face data generation unit includes a region determination unit configured to: determine the cheek region based on the eyebrow key points, eye key points, nose key points, mouth key points and facial contour key points; and determine the pupil region based on the eye key points.

[0023] Another aspect of this disclosure provides an electronic device including one or more processors and a storage device, wherein the storage device is used to store executable instructions that, when executed by the processor, implement the method described above.

[0024] Another aspect of this disclosure provides a computer-readable storage medium storing computer-executable instructions that, when executed, are used to implement the methods described above.

[0025] Another aspect of this disclosure provides a computer program that includes computer-executable instructions that, when executed, are used to implement the methods described above.

[0026] According to embodiments of this disclosure, when acquiring video data containing a face, light control information is simultaneously acquired. This light control information enables control of ambient light when the user inputs video data containing a face. By determining first change information related to the cheek area, it can be determined whether the video has relevant facial change features of ordinary people. Simultaneously, by determining second change information related to the pupil area, it can be determined whether the face in the video is the face video data that was recorded at that time. Based on the first change information, the second change information, and the light control information, the face recognition result is determined, which can effectively improve the accuracy of face recognition, increase the difficulty of cracking, and improve the security of the method. Attached Figure Description

[0027] The foregoing contents, as well as other objects, features, and advantages of this disclosure, will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0028] Figure 1 The diagram illustrates a system architecture for a face recognition method based on Euler image magnification that can be applied to embodiments of the present disclosure.

[0029] Figure 2 A flowchart illustrating a face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown schematically.

[0030] Figure 3 The flowchart illustrating the operation S220 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown in the schematic diagram.

[0031] Figure 4 The flowchart illustrating the operation S221 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown in the schematic diagram.

[0032] Figure 5 The flowchart illustrating the operation S2212 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown in the illustration.

[0033] Figure 6 The flowchart illustrating the operation S330 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown in the schematic diagram.

[0034] Figure 7 The flowchart illustrating the operation S230 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown in the schematic diagram.

[0035] Figure 8 The flowchart illustrating the operation S231 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown in the schematic diagram.

[0036] Figure 9 The flowchart illustrating the operation S240 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown in the schematic diagram.

[0037] Figure 10A This illustration schematically shows a process of generating baseband in a face recognition method based on Euler image magnification according to an embodiment of the present disclosure;

[0038] Figure 10B This illustration schematically shows a process for determining the cheek region in a face recognition method based on Euler image magnification according to an embodiment of the present disclosure;

[0039] Figure 10C This illustration schematically shows a process for determining the pupil region in a face recognition method based on Euler image magnification according to an embodiment of the present disclosure;

[0040] Figure 11 A block diagram of a face recognition device based on Euler image magnification according to an embodiment of the present disclosure is schematically shown; and

[0041] Figure 12 A block diagram of an electronic device according to an embodiment of the present disclosure is shown schematically. Detailed Implementation

[0042] The embodiments of the present disclosure will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the disclosure. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the present disclosure for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concepts of the present disclosure.

[0043] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. The terms “comprising,” “including,” etc., as used herein indicate the presence of features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0044] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0045] When using expressions such as "at least one of A, B, or C," it should generally be interpreted in accordance with the meaning commonly understood by those skilled in the art (e.g., "a system having at least one of A, B, or C" should include, but is not limited to, systems having A alone, having B alone, having C alone, having A and B, having A and C, having B and C, and / or having A, B, and C, etc.). The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, features defined with "first" or "second" may explicitly or implicitly include one or more features.

[0046] In the technical solution of this course, all user information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, data stored, data displayed) involved are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation access points for users to choose to authorize or refuse.

[0047] In this article, the term "light control information" refers to information on ambient light control during the acquisition of video data containing human faces. Light control information may include, for example, controlling the screen brightness of a display or the lighting equipment of the video acquisition device.

[0048] Eulerian Video Magnification (EVM) is an image processing technique used to enhance and amplify minute changes and motions in video.

[0049] In existing technologies, the biggest challenge posed to banking systems by identity authentication and authorization verification is verifying the authenticity of information, such as facial video data. To ensure the authenticity of the entered facial information, users typically need to have their facial video captured at a designated service location, and then have it verified by relevant staff. Alternatively, a collaborative online and human model can be used: application submitted online, facial data captured offline, then submitted for review online, and finally the progress can be tracked online, with branch staff primarily responsible for verifying identity. Another option is an interactive online model, where the customer completes the entire online application process, requiring interactive verification of facial and other identity information.

[0050] The aforementioned video data containing faces requires users to complete interactive tasks according to preset actions during the face registration process, which involves multiple verifications by the user. In addition, this type of verification method has low security and is easily cracked through related video methods, making it impossible to effectively determine the authenticity of the face.

[0051] To address the shortcomings of the prior art, embodiments of this disclosure provide a face recognition method, apparatus, electronic device, readable storage medium, and computer program product based on Euler image magnification. This allows users to seamlessly input face video data without following preset actions, while accurately determining the authenticity of video data containing faces, thereby improving data security and reliability.

[0052] The face recognition method based on Euler image magnification in this disclosure includes, but is not limited to: acquiring video data containing a face and lighting control information, wherein the lighting control information includes a time series associated with the video data; determining the cheek region and pupil region in the face data based on the video data; magnifying the change information of the face data using the Euler image magnification algorithm to determine first change information associated with the cheek region and second change information associated with the pupil region; and determining the face recognition result based on the first change information, the second change information, and the lighting control information.

[0053] According to embodiments of this disclosure, when acquiring video data containing a face, light control information is simultaneously acquired. This light control information enables control of ambient light when the user inputs video data containing a face. By determining first change information related to the cheek area, it can be determined whether the video has relevant facial change features of ordinary people. Simultaneously, by determining second change information related to the pupil area, it can be determined whether the face in the video is the face video data that was recorded at that time. Based on the first change information, the second change information, and the light control information, the face recognition result is determined, which can effectively improve the accuracy of face recognition, increase the difficulty of cracking, and improve the security of the method.

[0054] Figure 1 A schematic diagram illustrates a system architecture for a face recognition method based on Euler image magnification, to which embodiments of this disclosure can be applied. It should be noted that... Figure 1 The examples shown are merely examples of system architectures applicable to the embodiments of this disclosure, intended to help those skilled in the art understand the technical content of this disclosure. However, they do not imply that the embodiments of this disclosure cannot be used in other devices, systems, environments, or scenarios. It should be noted that the face recognition method based on Euler image magnification provided in the embodiments of this disclosure can be used in the fields of artificial intelligence, big data, and finance, particularly in the area of ​​big data. It can also be used in any field other than finance. The face recognition method and apparatus based on Euler image magnification provided in the embodiments of this disclosure do not limit the application field.

[0055] like Figure 1 As shown, an exemplary system architecture 100 for applying a face recognition method based on Euler image magnification may include terminal devices 101, 102, and 103, a network 104, and a server 105. Network 104 serves as the medium for providing a communication link between terminal devices 101, 102, and 103 and server 105. Network 104 may include various connection types, such as wired or wireless communication links or fiber optic cables, etc.

[0056] Users can use terminal devices 101, 102, and 103 to interact with server 105 via network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, and 103, such as email client applications, file processing applications, shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only).

[0057] Terminal devices 101, 102, and 103 can be various electronic devices with displays that support functions such as data input, file sending, data analysis, data processing, and web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0058] Server 105 can be a server providing various services, such as a backend management server (for example only) that supports data obtained by users using terminal devices 101, 102, and 103 or websites they browse. The backend management server can analyze and process received user requests and other data, and feed back the processing results (such as web pages, information, or data obtained or generated based on user requests) to the terminal devices. It can also analyze or process files sent by users and control the terminal devices based on the processing results, such as restricting access to the terminal devices.

[0059] It should be noted that the face recognition method based on Euler image magnification provided in the embodiments of this disclosure can generally be executed by terminal devices 101, 102, 103, and server 105. Correspondingly, the face recognition device based on Euler image magnification provided in the embodiments of this disclosure can generally be located in terminal devices 101, 102, 103, and server 105. The face recognition method based on Euler image magnification provided in the embodiments of this disclosure can also be executed by a terminal device different from terminal devices 101, 102, and 103, but capable of communicating with terminal devices 101, 102, and 103. Correspondingly, the face recognition device based on Euler image magnification provided in the embodiments of this disclosure can also be located in a terminal device different from terminal devices 101, 102, and 103, but capable of communicating with terminal devices 101, 102, and 103.

[0060] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0061] The following will be through Figures 2 to 10C The face recognition method based on Euler image magnification according to the embodiments of this disclosure will be described in detail.

[0062] Figure 2 A flowchart illustrating a face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown schematically.

[0063] like Figure 2 As shown, the process 200 of the face recognition method based on Euler image magnification in this embodiment includes operations S210 to S240.

[0064] In operation S210, video data containing a face and lighting control information are acquired, the lighting control information including a time series associated with the video data.

[0065] In embodiments of this disclosure, users can collect video data containing faces using devices such as mobile phones, computers, tablets, and cameras, and then upload the data to complete face recognition. To verify the authenticity of the faces in the uploaded video data, light control information is sent to the collection device when the user collects the video data, thereby controlling the light during the video collection process. The light control information includes a time series associated with the video data. This allows for correlation between the time series and the video data during face verification, facilitating subsequent verification of changes in the pupil area within the face data. For example, by using the time series, it can be determined whether the changes in the pupil area are consistent with the changes in the light control information. If they are consistent in the time series, it indicates that the face recognition result in the video data is genuine.

[0066] In some embodiments of this disclosure, when acquiring video data containing faces, the video data includes information about the face and other parts of the body, such as the body. Because the user may move or change their distance from the acquisition device during video data acquisition, the video frames containing faces may vary. To address this, the acquired video data can be preprocessed, for example, by deleting video frames that do not contain faces from the video sequence. This improves processing speed and reduces interference signals in subsequent data processing.

[0067] For example, the light control signal can control the acquisition light throughout the entire video data acquisition process, or it can be a control that varies over time. For instance, it could increase the light brightness for a period of time at the beginning of acquisition and decrease the light brightness for a period of time before the acquisition ends. Alternatively, the light brightness can vary during the video data acquisition process, where the light brightness is adjusted through light control information.

[0068] In operation S220, the cheek region and pupil region in the face data are determined based on the video data.

[0069] In the embodiments of this disclosure, different video frames in the video data may contain other data besides the face, such as limbs, torso, etc. In order to reduce the interference of other data, the video data can be processed, such as segmentation, alignment, etc., to determine the face data, and further information on the cheek area and pupil area in the face data can be obtained, effectively reducing other interference information and improving the accuracy of face recognition results.

[0070] In operation S230, the change information of the face data is magnified by the Euler image magnification algorithm to determine the first change information associated with the cheek region and the second change information associated with the pupil region.

[0071] In some embodiments of this disclosure, real human faces undergo subtle periodic changes in skin due to blood circulation. These changes, imperceptible to the naked eye, correspond to changes in heart rate and are difficult to imitate. By utilizing these changes appropriately, it is possible to effectively determine whether a face on a screen is real. To this end, this disclosure uses the Euler image magnification algorithm to magnify facial data, thereby accurately determining changes in the facial data. For example, it determines first change information associated with the cheek area and second change information associated with the pupil area. This reduces the possibility of errors in data judgment. Furthermore, during the judgment process, users do not need to record video data according to pre-set actions, making the operation simple and efficient, improving user experience and convenience.

[0072] In the embodiments of this disclosure, the periodic changes in blood circulation are independent of ambient light. Therefore, the change period can be fixed. If compromised, it can be achieved by forging a change rate identical to the fixed change period of an ordinary person. Therefore, relying solely on sensing the change period caused by blood circulation offers low security. In contrast, light control information can be adjusted through settings, such as personalized settings based on different users. The pupil undergoes neural reflex changes under different lighting conditions; for example, the pupil dilates in dim light and constricts in bright light. Since the light control information is adjustable, pupil changes can be controlled. When user-associated light control information is unavailable, mimicking pupil changes is extremely difficult. Therefore, the embodiments of this disclosure can effectively improve the accuracy and security of facial recognition.

[0073] In operation S240, the face recognition result is determined based on the first change information, the second change information, and the light control information.

[0074] In the embodiments of this disclosure, since the first change information is the periodic change in human skin due to blood circulation, changes imperceptible to the naked eye can be used to determine whether a face in the video data is real. Furthermore, using the second change information and lighting control information to further determine the face recognition result can effectively reduce interference from other information and increase the difficulty of cracking, thereby improving the security and accuracy of the face recognition result.

[0075] Figure 3 The flowchart illustrating the operation S220 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown.

[0076] like Figure 3 As shown, the face recognition method based on Euler image magnification in this embodiment of the present disclosure includes operations S221 to S222 in the operation S220 process.

[0077] In operation S221, facial data is generated based on video data.

[0078] In some embodiments of this disclosure, since the collected video data containing faces may contain redundant information, such as other interfering information besides faces, in order to reduce the interfering information, the video data is processed by segmentation and alignment, thereby generating face data containing less interfering information, thereby improving the accuracy of face recognition results.

[0079] In operation S222, the cheek area and pupil area are determined based on the facial data.

[0080] In some embodiments of this disclosure, facial data, such as the cheek area, contains abundant blood vessels. During blood circulation, the periodic changes in this area are more obvious, while other parts, such as the ears and nose, show weaker periodic changes. Change information can be extracted from the cheek area to improve the accuracy of the obtained change information.

[0081] Furthermore, the pupil region is highly sensitive to changes in light. In order to determine the changes associated with light control information, the pupil region in the face data is identified. In subsequent operations, it is determined whether the second change information associated with the pupil region is consistent with the change in light control information, thereby more accurately determining the authenticity of the face recognition result.

[0082] Figure 4 The flowchart illustrating the operation S221 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown.

[0083] like Figure 4 As shown, the face recognition method based on Euler image magnification in this embodiment of the present disclosure includes operations S2211 to S2212 in the operation S221 process.

[0084] In operation S2211, face segmentation is performed on the video data to extract the face region in each video frame.

[0085] In some embodiments of this disclosure, face segmentation of video data can be achieved, for example, by using machine learning methods to determine face regions from the data of each video frame. Specifically, face regions are located and extracted from each video frame in the video sequence. This helps ensure more accurate and consistent results in subsequent analysis and comparison, while improving the utilization rate of face information in the video sequence.

[0086] In operation S2212, the extracted face regions are aligned to generate face data.

[0087] According to embodiments of this disclosure, aligning the extracted face regions can ensure that the face has a consistent position and angle across different video frames, thereby facilitating subsequent extraction and comparison of face features.

[0088] Figure 5 The flowchart illustrating the operation S2212 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown.

[0089] like Figure 5 As shown, the face recognition method based on Euler image magnification in this embodiment of the present disclosure includes operations S310 to S330 in operation S2212.

[0090] In operation S310, facial key points are determined based on the face region.

[0091] For example, facial key points include: eyebrow key points, eye key points, nose key points, mouth key points, and facial contour key points. By using different facial key points, different areas of the face can be more accurately identified.

[0092] In operation S320, a geometric transformation is performed on the face region to place the facial key points in a preset fixed position.

[0093] For example, geometric transformations can be applied to different locations on a face to correct the face, thereby allowing for more accurate extraction of change information during information retrieval.

[0094] In operation S330, the face regions after geometric transformation in each video frame are aligned to generate face data.

[0095] By aligning the transformed face regions, it is easier to amplify changes in face data during subsequent operations, reduce interference, and improve the accuracy of face recognition.

[0096] Figure 6 The flowchart illustrating the operation S330 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown.

[0097] like Figure 6 As shown, the face recognition method based on Euler image magnification in this embodiment of the present disclosure includes operations S331 to S332 in the operation S330 process.

[0098] In operation S331, the cheek area is determined based on eyebrow key points, eye key points, nose key points, mouth key points, and facial contour key points.

[0099] In the embodiments of this disclosure, the cheek area is an irregular area, which requires joint determination based on other key points such as eyebrow key points, eye key points, nose key points, mouth key points and facial contour key points, thereby reducing interference information from other irrelevant areas and improving accuracy.

[0100] In operation S332, the pupil area is determined based on key points of the eye.

[0101] The pupil is located in the area of ​​the eye's key points; therefore, the pupil area is determined based on these key points.

[0102] Figure 7 The flowchart illustrating the operation S230 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown.

[0103] like Figure 7 As shown, the face recognition method based on Euler image magnification in this embodiment of the present disclosure includes operations S231 to S233 in operation S230.

[0104] In operation S231, spatial filtering and downsampling are performed on the cheek region and pupil region in the face data to generate the first baseband and the second baseband.

[0105] In some embodiments of this disclosure, the cheek region and the pupil region each comprise multiple pixel matrices. By performing spatial filtering on the cheek region and the pupil region, low-frequency information (e.g., smooth areas, slow changes) in the image can be preserved, while high-frequency noise or details in the image can be removed. Downsampling can reduce the image resolution and remove some pixels to reduce image details, facilitating the construction of a Gaussian pyramid. By performing spatial filtering and downsampling on the cheek region in the face data, a first baseband can be generated. By performing spatial filtering and downsampling on the pupil region in the face data, a second baseband can be generated.

[0106] In operation S232, a bandpass filter is used in the time domain to perform bandpass filtering on the first baseband and the second baseband to extract the first change sub-information and the second change sub-information.

[0107] For example, sub-information I(x, t) at any time t has:

[0108] I(x, t) = f(x + δ(t)), t > 0

[0109] I(x, 0) = f(x), t = 0

[0110] Where δ(t) represents the extracted sub-information of changes.

[0111] Based on the above, the first variation sub-information and the second variation sub-information can be extracted for the first baseband and the second baseband, respectively.

[0112] In operation S233, the first change sub-information and the second change sub-information are amplified to a set factor and merged with the first baseband and the second baseband respectively to determine the first change information and the second change information.

[0113] For example, in the steps above, the acquired change information is approximated using a first-order Taylor series:

[0114]

[0115] Then, the approximation sub-information is amplified by a factor of α, and then merged with the atomic information. The merged sub-information is then superimposed on the original image.

[0116]

[0117] Ideally, It is approximately equal to I(x, t) in the above text, and finally the change information is determined through analysis.

[0118] Specifically, the first variation sub-information is amplified to a set factor α1 and combined with the first baseband signal to determine the first variation information. The second variation sub-information is amplified to a set factor α2 and combined with the second baseband signal to determine the first variation information. For example, the set factors α1 and α2 can be adjusted according to actual needs.

[0119] Figure 8 The flowchart illustrating the operation S231 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown.

[0120] like Figure 8 As shown, the face recognition method based on Euler image magnification in this embodiment of the present disclosure includes operations S2311 to S2313 in the operation S231 process.

[0121] In operation S2311, the first pixel value matrix of the color channel in the cheek region and the second pixel matrix of the color channel in the pupil region are obtained.

[0122] In operation S2312, a low-pass filter is used to spatially filter the first pixel value matrix to obtain first data with different spatial frequencies, and a spatial filter is used to spatially filter the second pixel matrix to obtain second data with different spatial frequencies.

[0123] In operation S2313, the first baseband and the second baseband are obtained by smoothing and downsampling the first data and the second data respectively according to the Gaussian pyramid.

[0124] For example, the first baseband refers to the lowest frequency range associated with the first pixel value matrix at different spatial frequencies, and the second baseband refers to the lowest frequency range associated with the second pixel value matrix at different spatial frequencies.

[0125] The first and second basebands contain the overall structure and low-frequency details of the image, which can reduce interference information and effectively improve the accuracy of subsequent operations.

[0126] Figure 9 The flowchart illustrating the operation S240 of the face recognition method based on Euler image magnification according to an embodiment of the present disclosure is shown.

[0127] like Figure 9 As shown, the face recognition method based on Euler image magnification in this embodiment of the present disclosure includes operations S241 to S245 in operation S240.

[0128] In some embodiments of this disclosure, the first set threshold includes the actual facial color change cycle, which may be calculated in advance based on existing data.

[0129] The second set threshold includes the actual pupillary nerve reflex change cycle, which can be calculated based on existing data, for example.

[0130] In embodiments of this disclosure, the judgment model for determining the first change information may, for example, be pre-trained based on the calibrated color change cycles of real and fake faces.

[0131] The judgment model for determining the second change information can be, for example, pre-trained based on the calibrated real pupillary nerve reflex change cycle and the spurious pupillary nerve reflex change cycle.

[0132] In operation S241, a pre-trained judgment model is used to determine whether the first change information meets the first set threshold.

[0133] In operation S242, a pre-trained judgment model is used to determine whether the second change information meets the second set threshold.

[0134] In operation S243, the second change information is compared with the light control information to determine if there is any consistency in the change.

[0135] For example, the second change information is synchronized with the change of the light control information in the time series. In the second change information, at time t1, the pupil constricts and the light control information controls the light to increase. In the second change information, at time t2, the pupil increases and the light control information controls the light to decrease. This shows that the second change information and the light control information are consistent in their changes.

[0136] In operation S244, if the first change information satisfies the first preset threshold, the second change information satisfies the second preset threshold, and the second change information shows consistency with the light control information, then the face recognition result is confirmed as true. That is, operations S241, S242, and S243 all meet the conditions, and the face recognition result is true.

[0137] In operation S245, if the first change information does not meet the first preset threshold, the second change information does not meet the second preset threshold, or the second change information does not show consistency with the light control information, then the face recognition result is confirmed as false. That is, if any one of operations S241, S242, or S243 fails to meet the condition, the face recognition result is false.

[0138] Figure 10A The illustration shows a schematic diagram of the process of generating baseband in a face recognition method based on Euler image magnification according to an embodiment of the present disclosure. Figure 10B The illustration shows a schematic diagram of the process of determining the cheek region in a face recognition method based on Euler image magnification according to an embodiment of the present disclosure. Figure 10C The illustration shows a schematic diagram of the process of determining the pupil region in a face recognition method based on Euler image magnification according to an embodiment of the present disclosure.

[0139] like Figure 10A As shown in the diagram 400, which illustrates the processing of the cheek region, the first pixel value matrix 401 of the color channel in the cheek region is obtained. A low-pass filter is used to perform spatial filtering on the first pixel value matrix 401 to obtain first data (402, 403, 404) with different spatial frequencies. The first data is then smoothed and downsampled to obtain the first baseband 405.

[0140] like Figure 10B and Figure 10C As shown, segmenting and aligning the video data can generate 500 face data points. Then, as... Figure 10A As shown, the cheek region 501 is determined based on key points in the facial data, as follows. Figure 10B As shown, based on key points in the facial data, the pupil region 502 is determined, which facilitates the determination of the first change information associated with the cheek region and the second change information associated with the pupil region in subsequent operations.

[0141] According to embodiments of this disclosure, when acquiring video data containing a face, light control information is simultaneously acquired. This light control information enables control of ambient light when the user inputs video data containing a face. By determining first change information related to the cheek area, it can be determined whether the video has relevant facial change features of ordinary people. Simultaneously, by determining second change information related to the pupil area, it can be determined whether the face in the video is the face video data that was recorded at that time. Based on the first change information, the second change information, and the light control information, the face recognition result is determined, which can effectively improve the accuracy of face recognition, increase the difficulty of cracking, and improve the security of the method.

[0142] Another aspect of this disclosure provides a face recognition device based on Euler image magnification.

[0143] Figure 11 A block diagram of a face recognition device based on Euler image magnification according to an embodiment of the present disclosure is shown schematically.

[0144] like Figure 11 As shown, the face recognition device 600 based on Euler image magnification includes: an acquisition module 601, a first determination module 602, a second determination module 603, and a third determination module 604.

[0145] The acquisition module 601 is configured to acquire video data containing a face and lighting control information, wherein the lighting control information includes a time series associated with the video data. In one embodiment, the acquisition module 601 can be used to perform the operation S210 described above, which will not be repeated here.

[0146] The first determining module 602 is configured to determine the cheek region and pupil region in the face data based on the video data. In one embodiment, the first determining module 602 can be used to perform the operation S220 described above, which will not be repeated here.

[0147] The second determining module 603 is configured to amplify the change information of the face data using the Euler image magnification algorithm, and determine the first change information associated with the cheek region and the second change information associated with the pupil region. In one embodiment, the second determining module 603 can be used to perform the operation S230 described above, which will not be repeated here.

[0148] The third determining module 604 is configured to determine the face recognition result based on the first change information, the second change information, and the lighting control information. In one embodiment, the third determining module 604 can be used to perform the operation S240 described above, which will not be repeated here.

[0149] In some embodiments of this disclosure, the third determining module includes a result generating module, configured to: determine whether the first change information meets the first set threshold and whether the second change information meets the second set threshold using a pre-trained judgment model; compare whether the second change information and the light control information have change consistency; if the first change information meets the first set threshold, the second change information meets the second set threshold, and the second change information and the light control information have change consistency, then confirm that the face recognition result is true; otherwise, determine that the face recognition result is false.

[0150] In some embodiments of this disclosure, the second determining module includes a change information determining module configured to: perform spatial filtering and downsampling on the cheek region and pupil region in the face data to generate a first baseband and a second baseband; perform bandpass filtering on the first baseband and the second baseband in the time domain to extract first change sub-information and second change sub-information; amplify the first change sub-information and the second change sub-information to a set factor, and merge them with the first baseband and the second baseband respectively to determine the first change information and the second change information.

[0151] In some embodiments of this disclosure, the change information determination module includes a baseband generation module configured to: acquire a first pixel value matrix of the color channel in the cheek region and a second pixel matrix of the color channel in the pupil region; perform spatial filtering on the first pixel value matrix using a low-pass filter to obtain first data with different spatial frequencies, and perform spatial filtering on the second pixel matrix to obtain second data with different spatial frequencies; and perform smoothing and downsampling on the first data and the second data according to a Gaussian pyramid to obtain the first baseband and the second baseband respectively.

[0152] In some embodiments of this disclosure, the first determining module includes a first determining unit configured to: generate face data based on the video data; and determine the cheek region and pupil region based on the face data.

[0153] In some embodiments of this disclosure, the first determining unit includes a face data generation module, configured to: perform face segmentation on the video data, extract face regions from each video frame; and align the extracted face regions to generate face data.

[0154] In some embodiments of this disclosure, the face data generation module includes a face data generation unit configured to: determine facial key points based on the face region; perform geometric transformation on the face region to place the facial key points in a preset fixed position; and align the geometrically transformed face regions in each video frame to generate face data.

[0155] In some embodiments of this disclosure, the face data generation unit includes a region determination unit configured to: determine the cheek region based on the eyebrow key points, eye key points, nose key points, mouth key points and facial contour key points; and determine the pupil region based on the eye key points.

[0156] According to embodiments of this disclosure, any plurality of modules among the acquisition module 601, the first determining module 602, the second determining module 603, and the third determining module 604 can be combined into one module, or any one of these modules can be split into multiple modules. Alternatively, at least part of the functionality of one or more of these modules can be combined with at least part of the functionality of other modules and implemented in one module. According to embodiments of this disclosure, at least one of the acquisition module 601, the first determining module 602, the second determining module 603, and the third determining module 604 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging the circuitry, or implemented in any one of the three implementation methods of software, hardware, and firmware, or in a suitable combination of any of these. Alternatively, at least one of the acquisition module 601, the first determining module 602, the second determining module 603, and the third determining module 604 can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0157] Figure 12 A block diagram of an electronic device according to an embodiment of the present disclosure is shown schematically. Figure 12 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0158] like Figure 12 As shown, an electronic device 700 according to an embodiment of the present disclosure includes a processor 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage portion 708 into a random access memory (RAM) 703. The processor 701 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 701 may also include onboard memory for caching purposes. The processor 701 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0159] RAM 703 stores various programs and data required for the operation of electronic device 700. Processor 701, ROM 702, and RAM 703 are interconnected via bus 704. Processor 701 performs various operations of the method flow according to embodiments of the present disclosure by executing programs in ROM 702 and / or RAM 703. It should be noted that the programs may also be stored in one or more memories other than ROM 702 and RAM 703. Processor 701 may also perform various operations of the method flow according to embodiments of the present disclosure by executing programs stored in said one or more memories.

[0160] According to embodiments of this disclosure, the electronic device 700 may further include an input / output (I / O) interface 705, which is also connected to a bus 704. The electronic device 700 may also include one or more of the following components connected to the I / O interface 705: an input section 706 including a keyboard, mouse, etc.; an output section 707 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 710 as needed so that computer programs read from it can be installed into the storage section 708 as needed.

[0161] This disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs that, when executed, implement the method according to the embodiments of this disclosure.

[0162] According to embodiments of this disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium, such as, but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this disclosure, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this disclosure, the computer-readable storage medium may include ROM 702 and / or RAM 703 and / or one or more memories other than ROM 702 and RAM 703 described above.

[0163] Embodiments of this disclosure also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to cause the computer system to implement the methods provided in the embodiments of this disclosure.

[0164] When the computer program is executed by the processor 701, it performs the functions defined in the system / apparatus of this disclosure embodiments. According to embodiments of this disclosure, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0165] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 709, and / or installed from a removable medium 711. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0166] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 709, and / or installed from the removable medium 711. When the computer program is executed by the processor 701, it performs the functions defined in the system of this disclosure embodiment. According to embodiments of this disclosure, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0167] According to embodiments of this disclosure, program code for executing the computer programs provided in embodiments of this disclosure can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can execute entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0168] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0169] Those skilled in the art will understand that the features described in the various embodiments and / or claims of this disclosure can be combined or combined in various ways, even if such combinations or combinations are not explicitly described in this disclosure. In particular, the features described in the various embodiments and / or claims of this disclosure can be combined or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0170] The embodiments of this disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of this disclosure is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this disclosure, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A face recognition method based on Euler image magnification, wherein, include: Acquire video data containing a human face and lighting control information, wherein the lighting control information includes a time series associated with the video data; Based on the video data, determine the cheek region and pupil region in the face data; The Euler image magnification algorithm is used to amplify the change information of the facial data, determine the first change information associated with the cheek region and the second change information associated with the pupil region, including: spatial filtering and downsampling the cheek region and pupil region in the facial data to generate a first baseband and a second baseband; applying a bandpass filter to the first baseband and the second baseband in the time domain to extract the first change sub-information and the second change sub-information; amplifying the first change sub-information and the second change sub-information to a set factor, and merging them with the first baseband and the second baseband respectively to determine the first change information and the second change information; and Determining a face recognition result based on the first change information, the second change information, and the light control information includes: determining whether the first change information meets a first set threshold and whether the second change information meets a second set threshold using a pre-trained judgment model; comparing whether the second change information and the light control information have a change consistency; if the first change information meets the first set threshold, the second change information meets the second set threshold, and the second change information and the light control information have a change consistency, then the face recognition result is confirmed as true; otherwise, the face recognition result is determined as false.

2. The method according to claim 1, wherein, Spatial filtering and downsampling are performed on the cheek and pupil regions of the facial data to generate a first baseband and a second baseband, including: Obtain the first pixel value matrix of the color channel in the cheek region and the second pixel matrix of the color channel in the pupil region; A low-pass filter is used to spatially filter the first pixel value matrix to obtain first data with different spatial frequencies, and a low-pass filter is used to spatially filter the second pixel matrix to obtain second data with different spatial frequencies. The first baseband and the second baseband are obtained by smoothing and downsampling the first data and the second data respectively using the Gaussian pyramid.

3. The method according to claim 1, wherein, Based on the video data, the cheek region and pupil region in the facial data are determined, including: Based on the video data, generate facial data; and Based on the facial data, the cheek area and pupil area are determined.

4. The method according to claim 3, wherein, Based on the video data, facial data is generated, including: Perform face segmentation on the video data and extract the face region from each video frame; and The extracted face regions are aligned to generate face data.

5. The method according to claim 4, wherein, Aligning the extracted face regions to generate face data includes: Based on the facial region, determine the facial key points; Perform geometric transformations on the face region to position the facial key points at preset fixed locations; and The face regions after geometric transformation in each video frame are aligned to generate face data.

6. The method according to claim 5, wherein, The key facial features include: Key points for eyebrows, eyes, nose, mouth, and facial contours.

7. The method according to claim 6, wherein, Based on the facial data, the cheek region and pupil region are determined, including: The cheek area is determined based on the aforementioned key points of the eyebrows, eyes, nose, mouth, and facial contours; and The pupil region is determined based on the key points of the eye.

8. The method according to any one of claims 1 to 7, wherein, The first set threshold includes the cycle of real human face color change, and the second set threshold includes the cycle of real pupillary nerve reflex change.

9. A face recognition device based on Euler image magnification, wherein, include: The acquisition module is configured to acquire video data containing a face and lighting control information, wherein the lighting control information includes a time series associated with the video data; The first determining module is configured to determine the cheek region and pupil region in the face data based on the video data; The second determining module is configured to amplify the change information of the face data using the Euler image magnification algorithm, and determine the first change information associated with the cheek region and the second change information associated with the pupil region, including: performing spatial filtering and downsampling on the cheek region and pupil region in the face data to generate a first baseband and a second baseband; performing bandpass filtering on the first baseband and the second baseband in the time domain to extract the first change sub-information and the second change sub-information; amplifying the first change sub-information and the second change sub-information to a set factor, and merging them with the first baseband and the second baseband respectively to determine the first change information and the second change information; and The third determining module is configured to determine the face recognition result based on the first change information, the second change information, and the light control information, including: determining whether the first change information meets a first set threshold and whether the second change information meets a second set threshold using a pre-trained judgment model; comparing whether the second change information and the light control information have a change consistency; if the first change information meets the first set threshold, the second change information meets the second set threshold, and the second change information and the light control information have a change consistency, then the face recognition result is confirmed to be true; otherwise, the face recognition result is determined to be false.

10. An electronic device comprising one or more processors; A storage device for storing executable instructions, which, when executed by the processor, implement the method according to any one of claims 1 to 8.

11. A computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, implement the method according to any one of claims 1 to 8.

12. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 8.

Images