Face detection method, device, system, equipment and storage medium

Abstract

The present application discloses a face detection method, device, system, equipment and storage medium. The present application converts a target face image shot in a fixed shooting posture into target sequence information according to a fixed rule, and uses the target sequence information to represent the face information in the corresponding target face image. With the cooperation of short message communication, fixed shooting posture and fixed rules, even if the front-end device is far away from the control center, the face information of the person on the front-end device side can be sent to the control center. The control center saves the sequence information corresponding to the face images of multiple people. The face images are shot in a fixed shooting posture and converted into sequence information according to fixed rules. When the face image and the target face image contain the face information of the same person, the target sequence information and sequence information converted respectively are the same, so the identity of the person corresponding to the target sequence information can be determined by matching the target sequence information with the sequence information collected in advance.

Description

Technical Field

[0001] The present application relates to the field of face detection technology, and in particular to a face detection method, device, system, equipment and storage medium. Background Art

[0002] The face detection system is equipped with a control center and front-end equipment. The control center is connected to the front-end equipment through optical fiber. The front-end equipment collects face images and the control center manages the face database. The front-end equipment sends the face image to the control center through optical fiber. The control center matches the face image with the face database to determine the identity of the person corresponding to the front-end device.

[0003] When the control center and the front-end device are far apart, it is not convenient to connect the control center and the front-end device through optical fiber due to the high cost of optical fiber communication. Although the face database can be stored in the front-end device, the front-end device is responsible for face collection and identity authentication. However, the security capability of the front-end device is poor, and the memory of the front-end device is easily stolen, resulting in serious consequences of personnel information leakage. Therefore, the applicable scenarios of the existing face detection system are limited to the close setting of the control center and the front-end device, and the versatility is low. Summary of the invention

[0004] The present application provides a face detection method, apparatus, system, equipment and storage medium, which solves the problem in the prior art that a control center and front-end equipment cannot be set up at a long distance, and improves the versatility of the face detection system.

[0005] In a first aspect, the present application provides a face detection method, comprising:

[0006] Acquire a target facial image, and identify a corresponding number of feature points in the target facial image based on a preset number of feature points, wherein the target facial image is captured according to a set shooting posture;

[0007] Based on the Delaunay triangulation rule, the target face image is divided into a plurality of triangles formed by the feature points to obtain angles corresponding to the feature points;

[0008] Sorting the angles corresponding to the feature points based on a preset sorting rule to generate target sequence information corresponding to the target face image;

[0009] The target sequence information is sent to a high-orbit satellite via short message communication, so that the high-orbit satellite forwards the target sequence information to a control center, and the control center matches the target sequence information with sequence information in a face database to determine identity information corresponding to the target face image, wherein the sequence information is generated by the corresponding face image based on the number of feature points and the sorting rule, and the face image is captured according to the shooting posture.

[0010] In a second aspect, the present application provides a face detection device, comprising:

[0011] A first feature point determination module is configured to obtain a target face image, and identify a corresponding number of feature points in the target face image based on a preset number of feature points, wherein the target face image is taken according to a set shooting posture;

[0012] A first triangulation module is configured to divide the target face image into a plurality of triangles formed by the feature points based on a Delaunay triangulation rule, and obtain angles corresponding to the feature points;

[0013] A first sequence generating module is configured to sort the angles corresponding to the feature points based on a preset sorting rule to generate target sequence information corresponding to the target face image;

[0014] The face detection module is configured to send the target sequence information to a high-orbit satellite through short message communication, so that the high-orbit satellite forwards the target sequence information to a control center, and the control center matches the target sequence information with the sequence information in the face database to determine the identity information corresponding to the target face image, wherein the sequence information is generated by the corresponding face image based on the number of feature points and the sorting rule, and the face image is captured according to the shooting posture.

[0015] In a third aspect, the present application provides a face detection system, comprising: a front-end device, a high-orbit satellite, and a control center, wherein the front-end device is connected to the high-orbit satellite, and the high-orbit satellite is connected to the control center, wherein:

[0016] The front-end device is used to obtain a target facial image, identify a corresponding number of feature points in the target facial image based on a preset number of feature points, and the target facial image is obtained by shooting according to a set shooting posture; based on the Delaunay triangulation rule, divide the target facial image into a plurality of triangles formed by the feature points to obtain angles corresponding to the feature points; sort the angles corresponding to the feature points based on a preset sorting rule to generate target sequence information corresponding to the target facial image; and send the target sequence information to the high-orbit satellite through short message communication;

[0017] The high-orbit satellite is used to receive the target sequence information sent by the front-end device and send the target sequence information to the control center;

[0018] The control center is used to receive the target sequence information sent by the high-orbit satellite, and match the target sequence information with the sequence information in the face database to determine the identity information corresponding to the target face image, wherein the sequence information is generated by the corresponding face image based on the number of feature points and the sorting rule, and the face image is captured according to the shooting posture.

[0019] In a fourth aspect, the present application provides a face detection device, comprising:

[0020] One or more processors; a storage device storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the face detection method as described in the first aspect.

[0021] In a fifth aspect, the present application provides a storage medium comprising computer executable instructions, which, when executed by a computer processor, are used to perform the face detection method as described in the first aspect.

[0022] The present application converts the target face image shot in a fixed shooting posture into target sequence information according to a fixed rule, and uses the target sequence information to characterize the face information in the corresponding target face image. Since the data volume of the target face image far exceeds the data volume that can be carried by short message communication, after the target face image is converted into the corresponding target sequence information, its data volume can be reduced to less than the carrying range of short message communication, so the target sequence information can be transmitted over a long distance through short message communication. With the cooperation of short message communication, fixed shooting posture and fixed rules, even if the front-end device is far away from the control center, the face information of the person on the front-end device side can be sent to the control center. The control center saves the sequence information corresponding to the face images of multiple people, and the face image is also shot in a fixed shooting posture and converted into sequence information according to the fixed rule. When the face image and the target face image contain the face information of the same person, the target sequence information and sequence information converted respectively are the same, so the identity of the person corresponding to the target sequence information can be determined by matching the target sequence information with the sequence information collected in advance. Through the above technical means, the front-end device can transmit the facial information to the control center set up at an ultra-long distance. The control center can identify the facial information sent by the front-end device, solving the problem that the control center and the front-end device cannot be set up at a long distance. In addition, the front-end device only needs to convert the target facial image into the target sequence information, and there is no need to save the target facial image, which avoids the consequences of personnel information leakage caused by the theft of the front-end device and ensures the confidentiality of the front-end device. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a flow chart of a face detection method provided in an embodiment of the present application;

[0024] Figure 2 is a structural diagram of a face detection system provided in an embodiment of the present application;

[0025] Figure 3 is a schematic diagram of a target face image after Delaunay triangulation provided in an embodiment of the present application;

[0026] Figure 4 is a flowchart of generating target sequence information provided by an embodiment of the present application;

[0027] Figure 5 is a flow chart of an edge device detecting whether image data includes portrait information provided by an embodiment of the present application;

[0028] Figure 6 It is a flow chart of a front-end device generating sequence information and sending it to a control center provided by an embodiment of the present application;

[0029] Figure 7is a structural schematic diagram of a face detection device provided in an embodiment of the present application;

[0030] Figure 8 It is a structural schematic diagram of a face detection device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical scheme and advantages of the present application clearer, the specific embodiments of the present application are further described in detail below in conjunction with the accompanying drawings. It is understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for the convenience of description, only the part related to the present application but not all the contents are shown in the accompanying drawings. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow chart describes each operation (or step) as a sequential process, many of the operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of each operation can be rearranged. The process can be terminated when its operation is completed, but it can also have additional steps not included in the accompanying drawings. The process can correspond to a method, a function, a procedure, a subroutine, a subprogram, etc.

[0032] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0033] The face detection method provided in this embodiment can be executed by a face detection device, which can be implemented by software and / or hardware, and can be composed of two or more physical entities, or one physical entity. For example, the face detection device can be a front-end device equipped with an image acquisition device, or a processor of the front-end device.

[0034] The face detection device is installed with at least one type of operating system, and the face detection device can install at least one application based on the operating system, and the application can be an application that comes with the operating system, or an application downloaded from a third-party device or server. In this embodiment, the face detection device is installed with at least an application that can execute the face detection method, so the face detection device can also be the application itself.

[0035] For ease of understanding, this embodiment is described by taking a front-end device as an example of a main body for executing the face detection method.

[0036] In one embodiment, the face detection system includes a front-end device and a control center, the front-end device and the control center are connected by optical fiber, the front-end device is mainly used to collect facial information of personnel and transmit the facial information to the control center through optical fiber communication, and the control center is mainly used to determine the identity of the corresponding personnel according to the facial information collected by the front-end device. However, if the face detection system is to be set in an oil extraction plant in the desert, the front-end device is generally set in the desert, and the control center is generally set outside the desert. In this case, the cost of establishing a communication connection between the front-end device and the control center through optical fiber is too high, so the communication connection between the front-end device and the control center will not be established by pulling optical fiber. Although facial information can be stored in the front-end device, or the control center is also set in the desert, the security capabilities of the front-end device and the control center set in the desert are poor, and the storage of the front-end device or the control center is easily stolen, resulting in the leakage of personnel information. Therefore, the applicable scenarios of the existing face detection system are limited to areas with high security capabilities set at close range, and the versatility is poor.

[0037] To solve the above problem, this embodiment provides a face detection method to establish a communication connection between a remotely located control center and a front-end device through short message communication.

[0038] Figure 1 A flow chart of a face detection method provided by an embodiment of the present application is given. Figure 1 , the face detection method specifically includes:

[0039] S110, acquiring a target facial image, and identifying a corresponding number of feature points in the target facial image based on a preset number of feature points. The target facial image is obtained by photographing according to a set shooting posture.

[0040] The target face image can be understood as the face image of the person to be detected collected by the front-end device through the image acquisition device, which contains the face information of the corresponding person to be detected. Figure 2 is a structural diagram of a face detection system provided in an embodiment of the present application. Figure 2As shown, the face detection system includes a front-end device, a high-orbit satellite (GEO satellite) and a control center, wherein the high-orbit satellite is a satellite in the Beidou satellite navigation system. In this embodiment, the face information of the person to be detected on the front-end device side is transmitted to the control center through the short message communication technology of the Beidou satellite navigation and positioning system, and the control center determines the identity information of the person to be detected based on the face information. However, the amount of data that can be carried by the short message communication technology is 14000bit. If the face information of the person to be detected exceeds 14000bit, it cannot be transmitted to the control center through the short message communication technology. In general, the target face image can be regarded as a two-dimensional matrix, and its data amount far exceeds the amount of data that can be carried by the short message communication technology. Therefore, the method of performing face detection by the control center according to the target face image in the traditional technology is not applicable to the short message communication technology in this embodiment. In this regard, in order to realize the transmission of face information by the short board communication technology, the target face image is divided into multiple triangles by the Delaunay triangulation rule, and the angles of each triangle are sorted according to specific rules to generate target sequence information, so as to characterize the face information of the person to be detected by the target sequence information. The sequence is a one-dimensional array, and its data volume is smaller than the data volume that can be carried by the short message communication technology. Therefore, the sequence can be sent to the control center through the short message communication technology, and the control center matches the target sequence information with the pre-stored sequence information to determine the identity information of the person to be detected. Among them, the sequence information pre-stored by the control center is generated by the corresponding face image, and the face image is obtained by the corresponding person in advance according to the preset shooting posture. The conversion process of the face image into the corresponding sequence information is the same as the conversion process of the target face image into the corresponding target sequence information.

[0041] In this embodiment, the shooting posture of the person to be detected when shooting the target face image is the same as the shooting posture when shooting the face image in advance. If the same person to be detected shoots the target face image and the face image in different shooting postures, the triangles generated after Delaunay triangulation of the target face image and the face image will be different, which will lead to different target sequence information and sequence information, and face detection cannot be performed.

[0042] In this embodiment, before performing Delaunay triangulation, the feature points in the target facial image are first identified, that is, the pixel coordinates of the feature points in the target facial image are determined, so as to perform Delaunay triangulation on the target facial image based on the feature points. The number of feature points in the target facial image is pre-set, and the number of feature points will affect the distribution of the feature points, and the distribution structure of the feature points will affect the distribution of triangles. Therefore, in order to ensure the consistency of the target sequence information and the corresponding sequence information in the control center, the triangles corresponding to the target facial image and the triangles corresponding to the facial image should be the same, that is, the number of feature points corresponding to the target facial image is the same as the number of feature points corresponding to the facial image.

[0043] In one embodiment, the more feature points there are, the more accurate the face information represented by the triangular structure corresponding to the target face image is, that is, the higher the accuracy of the target sequence information is, the more accurate the face detection result is. Therefore, the number of feature points can be increased as much as possible within the data volume carrying range of short message communication, and the number of feature points is generally set to about 50.

[0044] S120 . Based on the Delaunay triangulation rule, divide the target face image into a plurality of triangles formed by feature points, and obtain angles corresponding to the feature points.

[0045] For example, the feature point is used as the corner point of a triangle by the Delaunay triangulation rule, and the three feature points are connected to obtain the corresponding triangle. The angle corresponding to the feature point is the angle in the triangle corresponding to the corner point. Figure 3 It is a schematic diagram of the target face image after Delaunay triangulation provided by the embodiment of the present application, which shows some triangular structures in the target face image. Points A to H are feature points in the target face image, and triangles ① to ⑧ are constructed in the target face image based on points A to H. Feature point A is the corner point of triangle ①, triangle ② and triangle ③, and its corresponding angles as the corner point of triangle ①, triangle ② and triangle ③ are a1, a2 and a3 respectively. The same is true for other feature points.

[0046] It should be noted that the target face image and the face image of the same person taken in a fixed shooting posture have the same feature points identified, and accordingly, the triangles generated by Delaunay triangulation based on the feature points are also the same. Therefore, for the same feature point, if the angle corresponding to the feature point at the left corner of the eye in the target face image and the feature point at the left corner of the eye in the corresponding face image are the same, the angle of the same feature point in the target face image and the face image can be compared to determine whether the target face image and the face image are of the same person.

[0047] S130 , sorting the angles corresponding to the feature points based on a preset sorting rule to generate target sequence information corresponding to the target face image.

[0048] Among them, the target sequence information refers to a sequence obtained by sorting the angles corresponding to all feature points in the target facial image, which includes all angles corresponding to all feature points in the target feature facial image. If the angles of each feature point in the target sequence information are the same as the angles of the corresponding feature points in a certain sequence information, it can be determined that the target facial image is the same as the facial image corresponding to the sequence information, that is, the identity information of the facial image is determined to be the identity information of the person to be detected. In this embodiment, in order to compare the angles of the same feature point in the target sequence information and the sequence information, both the feature points of the target facial image and the feature points of the facial image will be sorted according to a fixed sorting rule to ensure the consistency of the target sequence information and the corresponding sequence information, thereby realizing face detection.

[0049] In one embodiment, Figure 4 It is a flowchart of generating target sequence information provided by an embodiment of the present application.

[0050] like Figure 4 As shown, the generation of target sequence information includes S1301-S1302:

[0051] S1301. Sort the angles corresponding to the feature points according to a first sorting rule to determine the order of the angles corresponding to the feature points. The first sorting rule is from small to large or from large to small.

[0052] Exemplary, reference Figure 3 , the angles corresponding to feature point A are a1, a2 and a3. Assuming a1>a2>a3, according to the first sorting rule from large to small, the angle order of feature point A is a1a2a3. Similarly, the sorting process from small to large can be deduced.

[0053] It should be noted that when the angles corresponding to the feature points A in the facial image corresponding to the person to be detected are also sorted according to the first sorting rule of the angles corresponding to the feature points A in the target facial image, the angle order of the feature points A in the facial image is also a1a2a3.

[0054] S1302. According to the position information of each feature point in the target face image, the angle sequence of each feature point is sorted according to a second sorting rule to obtain target sequence information. The second sorting rule is any combination of from top to bottom or from bottom to top and from left to right or from left to right.

[0055] Exemplary, reference Figure 3, according to the ordinates of feature points A~H, sort feature points A~H from top to bottom and from left to right as ABCDEFGH, where feature point G and feature point H have the same ordinate, but feature point G is located on the left side of feature point H, so feature point G is placed in front of feature point H. Place the angle sequence corresponding to the feature point at the position of the corresponding feature point, such as the angle sequence a1a2a3 of feature point A is placed in front of the target sequence information. Reference Figure 3 , assuming that the angle order of feature point B is b1b2, the angle order of feature point C is c1c2c3c4, the angle order of feature point D is d2d1d3, the angle order of feature point E is e3e4e1e2e5e6, the angle order of feature point F is f1f2, the angle order of feature point G is g2g1, and the angle order of feature point H is h2h1, then according to the order of feature points A~J from top to bottom and from left to right ABCDEFGH, the angle order corresponding to the feature points is placed at the position of the corresponding feature points, and the target sequence information is a1a2a3b1b2c1c2c3c4d2d1d3e3e4e1e2e5e6f1f2g2g1h2h1. Similarly, it can be deduced that, according to the positions of the feature points in the target face image, the feature points are sorted from bottom to top and left to right, from top to bottom and right to left, from bottom to top and left to right, from right to left and top to bottom, from right to left and bottom to top, from left to right and top to bottom, or from left to right and bottom to top.

[0056] It should be noted that when the angle order corresponding to the feature points A~H in the facial image corresponding to the person to be detected is also sorted according to the second sorting rule of the angle order corresponding to the feature points A~H in the target facial image, the sequence information corresponding to the facial image is the same as the target sequence information, both of which are a1a2a3b1b2c1c2c3c4d2d1d3e3e4e1e2e5e6f1f2g2g1h2h1.

[0057] S140. Send the target sequence information to the high-orbit satellite through short message communication, so that the high-orbit satellite forwards the target sequence information to the control center, and the control center matches the target sequence information with the sequence information in the face database to determine the identity information corresponding to the target face image. The sequence information is generated by the corresponding face image based on the number of feature points and sorting rules, and the face image is captured according to the shooting posture.

[0058] Exemplarily, the target sequence information is sent to the high-orbit satellite of the Beidou satellite navigation and positioning system through the short message communication technology of the Beidou satellite navigation and positioning system, and the high-orbit satellite sends the target sequence information to the control center. The control center matches the target sequence information with the sequence information pre-stored in the face database, and determines that the identity information of the sequence information that is exactly the same as the target sequence information is the identity information of the person to be detected.

[0059] In one embodiment, reference Figure 2 , the face detection system is also provided with a ground center station (MCC), and the ground center station is connected to the high-orbit satellite. In this embodiment, the front-end device sends the target sequence information to the high-orbit satellite through short message communication, and the high-orbit satellite sends the target sequence information to the ground center station. The ground center station verifies whether the front-end device that sends the target sequence information is a legal sender, and sends the target sequence information to the control center through the high-orbit satellite when the front-end device is determined to be a legal sender. Exemplarily, the ground center station can verify whether the sender that sends the short message is a front-end device set for the face detection system. When it is determined that the sender is a front-end device set for the face detection system, it notifies the high-orbit satellite to send the short message to the control center. When it is determined that the sender is not a front-end device set for the face detection system, the short message is not sent to the control center to avoid the control center being interfered by information from other senders, thereby ensuring the reliability of face detection in the control center.

[0060] In one embodiment, reference Figure 2 , the face detection system also includes an edge server, which is connected to the front-end device, and the edge server can be set up in the core computer room. In this embodiment, the front-end device collects image data and sends the image data to the edge server. The edge server determines whether the image data includes the portrait information of the person to be detected, and sends a photo command to the front-end device when it is determined that the image data includes the portrait information of the person to be detected. After the front-end device receives the photo command sent by the edge server, it notifies the person to be detected to pose a set shooting posture according to the photo command, and shoots the target face image of the person to be detected. Exemplarily, the edge server can determine whether the object on the front-end device side is the person to be detected based on the image data. When the edge server determines that the object on the front-end device side is the person to be detected, it notifies the front-end device to collect the target face image of the person to be detected. When it is determined that the object on the front-end device side is not the person to be detected, it notifies the front-end device not to collect the target face image, thereby saving the power consumption of the front-end device in processing the target face image.

[0061] In another embodiment, reference Figure 2 , the face detection system also includes a 5G front-end device (5GCPE) and a 5G base station, and the front-end device is connected to the edge server through the 5G front-end device and the 5G base station. In this embodiment, the front-end device sends the image data to the 5G front-end device, the 5G front-end device sends the image data to the 5G base station, and the 5G base station sends the image data to the edge server. Exemplarily, arranging a 5G front-end device and a 5G base station between the edge server and the front-end device can effectively expand the data transmission distance between the edge server and the front-end device, and even if the edge server and the front-end device are set in different areas of the desert, the image data can be effectively transmitted to the edge server.

[0062] It should be noted that the communication distance between the 5G front-end device and the 5G base station is greater than the optical fiber communication distance but less than the short message communication distance of the Beidou satellite, and it cannot achieve ultra-long-distance transmission between the edge server and the front-end device, that is, the edge server is still set in the desert. However, because neither the edge server nor the front-end device in this embodiment will save images, there will be no problem of personnel information leakage.

[0063] In one embodiment, Figure 5 FIG. 1 is a flow chart of an edge device detecting whether image data includes portrait information provided by an embodiment of the present application. Figure 5 As shown, the step of the edge device detecting whether the image data includes portrait information specifically includes S210-S240:

[0064] S210. The edge server extracts multiple area pictures from the image data based on a preset first convolutional neural network.

[0065] S220. The edge server extracts corresponding feature information from the regional image based on a preset second convolutional neural network.

[0066] S230: Input the feature information into a preset support vector machine to obtain label information output by the support vector machine.

[0067] S240: Determine whether the image data includes portrait information according to the tag information.

[0068] Exemplarily, the first convolutional neural network extracts each region picture in the image data through a sliding window. Each region picture is preprocessed to adjust the size of the region picture to the size of the input picture in the second convolutional neural network. The preprocessed region picture is input into the second convolutional neural network, and the convolution kernel of the second convolutional neural network is convolved with the region picture to obtain feature information of the region picture. The feature information of each region picture is sequentially input into a support vector machine, and the support vector machine compares the region picture with sample images of various objects based on the feature information, determines the type of object contained in the region picture, and uses the object type as the label information of the corresponding input region picture. If the label information of the region picture is a person, it is determined that the image data includes portrait information. If the label information of the region picture is not a person, the feature information of the next region picture is input into the support vector machine. It can be understood that as long as the label information of one region picture is a person, the edge server can determine that the image data includes portrait information, otherwise it is determined that the image data does not include portrait information.

[0069] In one embodiment, a staff member can enter his or her own facial image in a fixed shooting posture at the control center side, and the control center will generate sequence information for the facial image according to a specific number of feature points and sorting rules, and associate the sequence information with the staff member's identity information and store it in a facial database.

[0070] In another embodiment, if the face detection system is used to verify whether the person to be detected has the pass qualifications, the control center can only save the sequence information corresponding to the staff who have the pass qualifications in the face database, so that when the face database is subsequently queried to find sequence information identical to the target sequence information, it can be determined that the person to be detected corresponding to the target sequence information has the pass qualifications.

[0071] In this embodiment, the sequence information of qualified personnel can also be obtained through the front-end device and sent to the control center through the short message communication of the Beidou satellite, and the control center saves the sequence information to the face database. Figure 6 This is a flow chart of the front-end device generating sequence information and sending it to the control center provided in the embodiment of the present application. Figure 6 As shown, the steps of generating sequence information and sending it to the control center by the front-end device specifically include S310-S340:

[0072] S310: Acquire at least one facial image, and identify feature points in the corresponding facial image based on the number of feature points.

[0073] S320, based on the Delaunay triangulation rule, divide the face image into a plurality of triangles formed by corresponding feature points, and obtain angles corresponding to the feature points in the face image.

[0074] S330 , sorting the angles corresponding to the feature points in the face image based on a sorting rule to generate sequence information corresponding to the face image.

[0075] S340. Send the sequence information to the high-orbit satellite via short message communication, so that the high-orbit satellite forwards the sequence information to the control center, and the control center stores the sequence information in the face database.

[0076] Exemplarily, a qualified staff member enters his / her own facial image in a fixed shooting posture on the front-end device side. The front-end device generates sequence information based on the specific number of feature points and sorting rules of the facial image, and sends the sequence information to the high-orbit satellite through the short message communication technology of the Beidou satellite. The high-orbit satellite forwards the sequence information to the control center, and the control center stores the sequence information in the face database. It can be understood that the implementation method of steps S310-S340 is similar to the implementation method of steps S110-S140. For details, please refer to steps S110-S140, which will not be repeated here.

[0077] In summary, the face detection method provided by the embodiment of the present application converts the target face image shot in a fixed shooting posture into target sequence information according to a fixed rule, and uses the target sequence information to characterize the face information in the corresponding target face image. Since the data volume of the target face image far exceeds the data volume that can be carried by short message communication, after the target face image is converted into the corresponding target sequence information, its data volume can be reduced to less than the carrying range of short message communication, so the target sequence information can be transmitted over a long distance through short message communication. With the cooperation of short message communication, fixed shooting posture and fixed rules, even if the front-end device is far away from the control center, the face information of the person on the front-end device side can be sent to the control center. The control center saves the sequence information corresponding to the face images of multiple persons, and the face image is also shot in a fixed shooting posture and converted into sequence information according to the fixed rule. When the face image and the target face image contain the face information of the same person, the target sequence information and sequence information converted respectively are the same, so the identity of the person corresponding to the target sequence information can be determined by matching the target sequence information with the sequence information collected in advance. Through the above technical means, the front-end device can transmit the facial information to the control center set up at an ultra-long distance. The control center can identify the facial information sent by the front-end device, solving the problem that the control center and the front-end device cannot be set up at a long distance. In addition, the front-end device only needs to convert the target facial image into the target sequence information, and there is no need to save the target facial image, which avoids the consequences of personnel information leakage caused by the theft of the front-end device and ensures the confidentiality of the front-end device.

[0078] Based on the above embodiments, Figure 7 A schematic diagram of the structure of a face detection device provided in an embodiment of the present application. Figure 7 The face detection device provided in this embodiment specifically includes: a first feature point determination module 41, a first triangulation module 42, a first sequence generation module 43 and a face detection module 44.

[0079] The first feature point determination module is configured to obtain a target face image, identify a corresponding number of feature points in the target face image based on a preset number of feature points, and the target face image is obtained by shooting according to a set shooting posture;

[0080] A first triangulation module is configured to divide the target face image into a plurality of triangles formed by feature points based on a Delaunay triangulation rule, and obtain angles corresponding to the feature points;

[0081] A first sequence generation module is configured to sort the angles corresponding to the feature points based on a preset sorting rule to generate target sequence information corresponding to the target face image;

[0082] The face detection module is configured to send the target sequence information to the high-orbit satellite through short message communication, so that the high-orbit satellite forwards the target sequence information to the control center, and the control center matches the target sequence information with the sequence information in the face database to determine the identity information corresponding to the target face image. The sequence information is generated by the corresponding face image based on the number of feature points and sorting rules, and the face image is captured according to the shooting posture.

[0083] On the basis of the above embodiment, the first sequence generation module includes: a first sorting unit, configured to sort the angles corresponding to the feature points according to a first sorting rule, and determine the angle order corresponding to the feature points, the first sorting rule being from small to large or from large to small; a second sorting unit, configured to sort the angle order of each feature point according to the second sorting rule according to the position information of each feature point in the target face image, and obtain target sequence information, the second sorting rule being any combination of from top to bottom or from bottom to top and from left to right or from left to right.

[0084] Based on the above embodiment, the face detection module includes: a short message transmission unit, which is configured to send the target sequence information to the high-orbit satellite through short message communication, so that the high-orbit satellite sends the target sequence information to the ground central station, and the ground central station verifies whether the front-end device sending the target sequence information is a legal sender, and sends the target sequence information to the control center via the high-orbit satellite when it is determined that the front-end device is a legal sender.

[0085] Based on the above embodiment, the face detection device also includes: a portrait determination module, which is configured to collect image data and send the image data to an edge server so that the edge server determines whether the image data includes portrait information, and sends a photo instruction to the front-end device when it is determined that the image data includes portrait information; a target image shooting module, which is configured to receive a photo shooting instruction sent by the edge server and shoot a target face image according to the photo shooting instruction.

[0086] Based on the above embodiment, the portrait determination module includes: a 5G communication unit, configured to send image data to a 5G front-end device, so that the 5G front-end device sends the image data to a 5G base station, and the 5G base station sends the image data to an edge server.

[0087] Based on the above embodiment, the portrait determination module includes: a portrait determination unit, configured to send the image data to the edge server, so that the edge server extracts multiple area pictures from the image data based on a preset first convolutional neural network, and extracts corresponding feature information from the area pictures based on a preset second convolutional neural network, inputs the feature information into a preset support vector machine, obtains label information output by the support vector machine, and determines whether the image data includes portrait information according to the label information.

[0088] On the basis of the above embodiment, the face detection device includes: a second feature point determination module, configured to obtain at least one face image, and identify feature points in the corresponding face image based on the number of feature points; a second triangulation module, configured to divide the face image into multiple triangles composed of corresponding feature points based on the Delaunay triangulation rule, and obtain the angles corresponding to the feature points in the face image; a second sequence generation module, configured to sort the angles corresponding to the feature points in the face image based on the sorting rule, and generate sequence information corresponding to the face image; a sequence information storage module, configured to send the sequence information to the high-orbit satellite through short message communication, so that the high-orbit satellite forwards the sequence information to the control center, and the control center stores the sequence information in the face database.

[0089] As described above, the face detection device provided by the embodiment of the present application converts the target face image shot in a fixed shooting posture into target sequence information according to a fixed rule, and uses the target sequence information to characterize the face information in the corresponding target face image. Since the data volume of the target face image far exceeds the data volume that can be carried by short message communication, after the target face image is converted into the corresponding target sequence information, its data volume can be reduced to less than the carrying range of short message communication, so the target sequence information can be transmitted over a long distance through short message communication. With the cooperation of short message communication, fixed shooting posture and fixed rules, even if the front-end device is far away from the control center, the face information of the person on the front-end device side can be sent to the control center. The control center saves the sequence information corresponding to the face images of multiple persons, and the face image is also shot in a fixed shooting posture and converted into sequence information according to the fixed rule. When the face image and the target face image contain the face information of the same person, the target sequence information and sequence information converted respectively are the same, so the identity of the person corresponding to the target sequence information can be determined by matching the target sequence information with the sequence information collected in advance. Through the above technical means, the front-end device can transmit the facial information to the control center set up at an ultra-long distance. The control center can identify the facial information sent by the front-end device, solving the problem that the control center and the front-end device cannot be set up at a long distance. In addition, the front-end device only needs to convert the target facial image into the target sequence information, and there is no need to save the target facial image, which avoids the consequences of personnel information leakage caused by the theft of the front-end device and ensures the confidentiality of the front-end device.

[0090] The face detection device provided in the embodiment of the present application can be used to execute the face detection method provided in the above embodiment, and has corresponding functions and beneficial effects.

[0091] Figure 8 is a structural diagram of a face detection device provided in an embodiment of the present application, with reference to Figure 8 The face detection device includes: a processor 51, a memory 52, a communication device 53, an input device 54 and an output device 55. The number of processors 51 in the face detection device can be one or more, and the number of memories 52 in the face detection device can be one or more. The processor 51, the memory 52, the communication device 53, the input device 54 and the output device 55 of the face detection device can be connected through a bus or other methods.

[0092] The memory 52, as a computer-readable storage medium, can be used to store software programs, computer executable programs and modules, such as program instructions / modules corresponding to the face detection method of any embodiment of the present application (for example, the first feature point determination module 41, the first triangulation module 42, the first sequence generation module 43 and the face detection module 44 in the face detection device). The memory 52 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the device, etc. In addition, the memory 52 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network and a combination thereof.

[0093] The communication device 53 is used for data transmission.

[0094] The processor 51 executes various functional applications and data processing of the device by running the software programs, instructions and modules stored in the memory 52, that is, implements the above-mentioned face detection method.

[0095] The input device 54 may be used to receive input digital or character information and generate key signal input related to user settings and function control of the device. The output device 55 may include a display device such as a display screen.

[0096] The face detection device provided above can be used to execute the face detection method provided in the above embodiment, and has corresponding functions and beneficial effects.

[0097] The embodiment of the present application also provides a storage medium containing computer executable instructions, which are used to execute a face detection method when executed by a computer processor. The face detection method includes: obtaining a target face image, identifying a corresponding number of feature points in the target face image based on a preset number of feature points, and the target face image is obtained by shooting according to a set shooting posture; dividing the target face image into a plurality of triangles composed of feature points based on a Delaunay triangulation rule, and obtaining angles corresponding to the feature points; sorting the angles corresponding to the feature points based on a preset sorting rule, and generating target sequence information corresponding to the target face image; sending the target sequence information to a high-orbit satellite through short message communication, so that the high-orbit satellite forwards the target sequence information to a control center, and the control center matches the target sequence information with the sequence information in a face database to determine the identity information corresponding to the target face image, and the sequence information is generated by the corresponding face image based on the number of feature points and the sorting rule, and the face image is obtained by shooting according to the shooting posture.

[0098] Storage medium - any of various types of memory devices or storage devices. The term "storage medium" is intended to include: installation media, such as CD-ROM, floppy disk or tape device; computer system memory or random access memory, such as DRAM, DDR RAM, SRAM, EDO RAM, Rambus RAM, etc.; non-volatile memory, such as flash memory, magnetic media (such as hard disk or optical storage); registers or other similar types of memory elements, etc. Storage media may also include other types of memory or combinations thereof. In addition, the storage medium may be located in the first computer system in which the program is executed, or may be located in a different second computer system, which is connected to the first computer system via a network (such as the Internet). The second computer system can provide program instructions to the first computer for execution. The term "storage medium" may include two or more storage media residing in different locations (for example, in different computer systems connected by a network). The storage medium may store program instructions (for example, embodied as a computer program) that can be executed by one or more processors.

[0099] Of course, the storage medium containing computer executable instructions provided in an embodiment of the present application, whose computer executable instructions are not limited to the above face detection method, can also execute related operations in the face detection method provided in any embodiment of the present application.

[0100] The face detection apparatus, face detection system, storage medium and face detection device provided in the above embodiments can execute the face detection method provided in any embodiment of the present application. For technical details not described in detail in the above embodiments, please refer to the face detection method provided in any embodiment of the present application.

[0101] The above are only preferred embodiments of the present application and the technical principles used. The present application is not limited to the specific embodiments herein, and various obvious changes, readjustments and substitutions that can be made by those skilled in the art will not deviate from the protection scope of the present application. Therefore, although the present application is described in more detail through the above embodiments, the present application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present application, and the scope of the present application is determined by the scope of the claims.

Claims

1. A face detection method, characterized in that: include: Collect image data, and send the image data to an edge server so that the edge server determines whether the image data includes portrait information, and sends a photo-taking instruction to a front-end device when it is determined that the image data includes portrait information, wherein sending the image data to the edge server so that the edge server determines whether the image data includes portrait information includes: sending the image data to the edge server so that the edge server extracts a plurality of area pictures from the image data based on a preset first convolutional neural network, extracts corresponding feature information from the area pictures based on a preset second convolutional neural network, inputs the feature information into a preset support vector machine, obtains label information output by the support vector machine, and determines whether the image data includes portrait information according to the label information; Receive a photographing instruction sent by the edge server, and photograph a target face image according to the photographing instruction; Acquire the target face image, and identify a corresponding number of feature points in the target face image based on a preset number of feature points, wherein the target face image is obtained by shooting according to a set shooting posture; Based on the Delaunay triangulation rule, the target face image is divided into a plurality of triangles formed by the feature points to obtain angles corresponding to the feature points; Sorting the angles corresponding to the feature points based on a preset sorting rule to generate target sequence information corresponding to the target face image; The target sequence information is sent to a high-orbit satellite via short message communication, so that the high-orbit satellite forwards the target sequence information to a control center, and the control center matches the target sequence information with the sequence information in a face database to determine the identity information corresponding to the target face image, wherein the sequence information is generated by the corresponding face image based on the number of feature points and the sorting rule, and the face image is captured according to the shooting posture, wherein the target sequence information is sent to a high-orbit satellite via short message communication, so that the high-orbit satellite forwards the target sequence information to the control center, including: sending the target sequence information to a high-orbit satellite via short message communication, so that the high-orbit satellite sends the target sequence information to a ground center station, and the ground center station verifies whether a front-end device that sends the target sequence information is a legitimate sender, and sends the target sequence information to the control center via the high-orbit satellite when it is determined that the front-end device is a legitimate sender.

2. The face detection method according to claim 1, characterized in that: The step of sorting the angles corresponding to the feature points based on a preset sorting rule to generate target sequence information corresponding to the target face image includes: Sorting the angles corresponding to the feature points according to a first sorting rule to determine the order of the angles corresponding to the feature points, wherein the first sorting rule is from small to large or from large to small; According to the position information of each feature point in the target face image, the angle order of each feature point is sorted according to a second sorting rule to obtain the target sequence information, and the second sorting rule is any combination of from top to bottom or from bottom to top and from left to right or from left to right.

3. The face detection method according to claim 1, characterized in that: The sending the image data to the edge server comprises: The image data is sent to a 5G front-end device, so that the 5G front-end device sends the image data to a 5G base station, and the 5G base station sends the image data to the edge server.

4. The face detection method according to claim 1, characterized in that: The method further comprises: Acquire at least one of the facial images, and identify feature points in the corresponding facial image based on the number of feature points; Based on the Delaunay triangulation rule, the face image is divided into a plurality of triangles formed by corresponding feature points, and the angles corresponding to the feature points in the face image are obtained; Sort the angles corresponding to the feature points in the face image based on the sorting rule to generate sequence information corresponding to the face image; The sequence information is sent to the high-orbit satellite via short message communication, so that the high-orbit satellite forwards the sequence information to the control center, and the control center stores the sequence information in the face database.

5. A face detection device, characterized in that: include: A portrait determination module is configured to collect image data and send the image data to an edge server so that the edge server determines whether the image data includes portrait information, and sends a photo-taking instruction to the front-end device when it is determined that the image data includes portrait information. Specifically, the module is configured to send the image data to the edge server so that the edge server extracts a plurality of area pictures from the image data based on a preset first convolutional neural network, extracts corresponding feature information from the area pictures based on a preset second convolutional neural network, inputs the feature information into a preset support vector machine, obtains label information output by the support vector machine, and determines whether the image data includes portrait information according to the label information; The target image shooting module is configured to receive a shooting instruction sent by the edge server and shoot a target face image according to the shooting instruction; A first feature point determination module is configured to obtain the target face image, and identify a corresponding number of feature points in the target face image based on a preset number of feature points, wherein the target face image is taken according to a set shooting posture; A first triangulation module is configured to divide the target face image into a plurality of triangles formed by the feature points based on a Delaunay triangulation rule, and obtain angles corresponding to the feature points; A first sequence generating module is configured to sort the angles corresponding to the feature points based on a preset sorting rule to generate target sequence information corresponding to the target face image; The face detection module is configured to send the target sequence information to a high-orbit satellite through short message communication, so that the high-orbit satellite forwards the target sequence information to a control center, and the control center matches the target sequence information with the sequence information in a face database to determine the identity information corresponding to the target face image, wherein the sequence information is generated by the corresponding face image based on the number of feature points and the sorting rule, and the face image is taken according to the shooting posture. The module is specifically configured to send the target sequence information to a high-orbit satellite through short message communication, so that the high-orbit satellite sends the target sequence information to a ground center station, and the ground center station verifies whether the front-end device that sends the target sequence information is a legal sender, and sends the target sequence information to the control center through the high-orbit satellite when it is determined that the front-end device is a legal sender.

6. A face detection system, characterized in that: include: A front-end device, a high-orbit satellite, a control center, an edge server and a ground central station, wherein the front-end device is connected to the high-orbit satellite and the edge server, and the high-orbit satellite is connected to the control center and the ground central station, wherein: The front-end device is used to collect image data and send the image data to the edge server; receive the photo-taking instruction sent by the edge server, and take a target face image according to the photo-taking instruction; The edge server is used to receive the image data sent by the front-end device, and determine whether the image data includes portrait information, and send a photo-taking instruction to the front-end device when it is determined that the image data includes portrait information, wherein the receiving of the image data sent by the front-end device and determining whether the image data includes portrait information includes: receiving the image data sent by the front-end device, extracting multiple area pictures from the image data based on a preset first convolutional neural network, and extracting corresponding feature information from the area pictures based on a preset second convolutional neural network, inputting the feature information into a preset support vector machine, obtaining label information output by the support vector machine, and determining whether the image data includes portrait information according to the label information; The front-end device is used to obtain the target face image, identify a corresponding number of feature points in the target face image based on a preset number of feature points, and the target face image is obtained by shooting according to a set shooting posture; based on the Delaunay triangulation rule, divide the target face image into a plurality of triangles formed by the feature points to obtain angles corresponding to the feature points; sort the angles corresponding to the feature points based on a preset sorting rule to generate target sequence information corresponding to the target face image; and send the target sequence information to the high-orbit satellite through short message communication; The high-orbit satellite is used to receive the target sequence information sent by the front-end device, send the target sequence information to the ground central station, and send the target sequence information to the control center; The ground central station is used to receive the target sequence information sent by the high-orbit satellite, and verify whether the front-end device sending the target sequence information is a legal sending end, and send the target sequence information to the control center via the high-orbit satellite when it is determined that the front-end device is a legal sending end; The control center is used to receive the target sequence information sent by the high-orbit satellite, and match the target sequence information with the sequence information in the face database to determine the identity information corresponding to the target face image, wherein the sequence information is generated by the corresponding face image based on the number of feature points and the sorting rule, and the face image is captured according to the shooting posture.

7. The face detection system according to claim 6, characterized in that: It includes a 5G front-end device and a 5G base station, the front-end device is connected to the 5G front-end device, the 5G front-end device is connected to the 5G base station, and the 5G base station is connected to the edge server, wherein: The front-end device is used to send the image data to the 5G front-end device; The 5G front-end device is used to receive the image data sent by the front-end device and send the image data to the 5G base station; The 5G base station is used to receive the image data sent by the 5G front-end device and send the image data to the edge server; The edge server is used to receive the image data sent by the 5G base station and determine whether the image data includes portrait information.

8. A face detection device, characterized in that: include: one or more processors; A storage device stores one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement the face detection method as described in any one of claims 1-4.

9. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to execute the face detection method as described in any one of claims 1 to 4 when executed by a computer processor.

Images