A face recognition method and system

Abstract

The present invention discloses a face authentication method, comprising the following steps: registering face storage features; setting storage feature tags and storage feature sources of face storage features; obtaining face detection features; setting detection feature tags and detection feature sources of face detection features; comparing storage feature sources and detection feature sources, and comparing storage feature tags and detection feature tags; if the comparison is successful, comparing face storage features and face detection features; if the face storage features and face detection features are successfully compared, face authentication is passed. The face authentication method disclosed by the present invention can improve the compatibility of face recognition products and is applicable to various occasions with different environments.

Description

Technical Field

[0001] The present invention relates to the field of face recognition technology, and more specifically, to a face authentication method and system. Background Art

[0002] Facial recognition technology is a type of biometric technology. It is a biometric recognition technology that automatically performs identity authentication based on the unique physiological characteristics of the face. It has the advantages of non-contact data collection, convenience, strong interactivity, and easy subsequent tracing.

[0003] Existing face recognition has a wide range of applications. For example, common access control systems can also use face recognition systems, that is, collect face data through a camera, and compare the collected face data with pre-stored authentication data. If the comparison is successful, face recognition is successful. If the comparison fails, face recognition fails.

[0004] The cameras currently used are generally visible light cameras and near-infrared cameras. However, the recognition devices corresponding to visible light cameras have the following problems: 1. They use visible light sources for identification and cannot be used in dark environments, and require supplementary light at night. 2. It is easy to use flat photos or other non-real faces for false authentication, which cannot effectively prevent counterfeiting. The recognition devices corresponding to near-infrared cameras also have the following problems: 1. They use short-wave light sources in the near-infrared band and rely on reflected light sources for imaging, and cannot be used in direct sunlight environments; 2. Although they can effectively prevent counterfeiting, facial features must be registered on the terminal device, which is not convenient enough.

[0005] Therefore, a face authentication method and system are designed to improve the compatibility of face recognition products, which are suitable for various occasions with different environments and provide the industry with more convenient and accurate face recognition technology. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide a face authentication method, improve the compatibility of face recognition products, and be applicable to various occasions with different environments.

[0007] To achieve the above object, the technical solution of the present invention is as follows:

[0008] A face recognition method comprises the following steps:

[0009] Register face storage features;

[0010] Setting a storage feature tag and a storage feature source of the face storage feature;

[0011] Get face detection features;

[0012] Setting a detection feature tag and a detection feature source of the face detection feature;

[0013] Comparing the stored feature source with the detected feature source, and comparing the stored feature tag with the detected feature tag;

[0014] If the comparison is successful, then comparing the stored face features with the face detection features;

[0015] If the facial storage features and the facial detection features are successfully compared, the facial authentication is passed;

[0016] Among them, the storage feature mark and the detection feature mark include a visible light mark and a near infrared mark, and the storage feature source and the detection feature source include a device source and a photo source.

[0017] Preferably, the setting of the storage feature mark and storage feature source of the face storage feature is specifically:

[0018] Setting the storage feature source;

[0019] If the storage feature source is the device source, setting the storage feature marker to the visible light marker and the near infrared marker;

[0020] If the storage feature source is the photo source, the storage feature mark is set to the visible light mark.

[0021] Preferably, the comparison of the storage feature source and the detection feature source, and the comparison of the storage feature mark and the detection feature mark; if the comparison is successful, the comparison of the face storage feature and the face detection feature is specifically:

[0022] comparing the stored feature source and the detected feature source;

[0023] If the comparison is successful, then comparing the stored characteristic mark and the detected characteristic mark;

[0024] If the stored feature mark and the detected feature mark are successfully compared, then the stored face feature and the detected face feature are compared.

[0025] Preferably, the comparison of the storage feature source and the detection feature source, and the comparison of the storage feature mark and the detection feature mark; if the comparison is successful, the comparison of the face storage feature and the face detection feature is specifically:

[0026] comparing the stored characteristic mark and the detected characteristic mark;

[0027] If the comparison is successful, then comparing the stored feature source and the detected feature source;

[0028] If the storage feature source and the detection feature source are successfully compared, the face storage feature and the face detection feature are compared.

[0029] Preferably, before comparing the storage feature source with the detection feature source, and comparing the storage feature mark with the detection feature mark, the method further includes:

[0030] Determining the detection feature mark;

[0031] If it is a visible light mark, near-infrared fill light is performed to reacquire the face detection feature, and the detection feature mark and the detection feature source of the face detection feature are updated.

[0032] Preferably, before comparing the storage feature source with the detection feature source, and comparing the storage feature mark with the detection feature mark, the method further includes:

[0033] Determining the detection feature mark;

[0034] If it is the near-infrared marker, visible light fill light is performed to reacquire the face detection feature, and the detection feature marker and the detection feature source of the face detection feature are updated.

[0035] A face recognition system, comprising:

[0036] A registration module, used to register face storage features;

[0037] A setting module connected to the registration module, used for setting the storage feature mark and storage feature source of the face storage feature;

[0038] An acquisition module connected to the setting module and used to acquire face detection features, and the setting module is also used to set detection feature tags and detection feature sources of the face detection features;

[0039] A comparison module connected to the setting module, used for comparing the storage feature source with the detection feature source, and for comparing the storage feature mark with the detection feature mark;

[0040] The comparison module is also used to compare the face storage feature and the face detection feature;

[0041] The authentication module connected to the comparison module is used to pass face authentication.

[0042] Preferably, it also includes:

[0043] A registration storage module connected between the registration module and the setting module, the registration storage module is used to store the face storage features;

[0044] An acquisition storage module connected between the acquisition module and the setting module, the acquisition storage module is used to store the face detection features.

[0045] Preferably, the acquisition module includes a visible light camera module and a near infrared camera module.

[0046] Preferably, it also includes:

[0047] The judging module connected to the acquiring module is used to judge the source of the detection feature;

[0048] A visible light fill light module connected to the visible light camera module;

[0049] A near-infrared fill light module connected to the near-infrared camera module;

[0050] If the judging module judges that it is the visible light mark, the near infrared fill light module is started, the acquiring module reacquires the face detection feature, and the setting module updates the detection feature mark and the detection feature source of the face detection feature;

[0051] If the judging module judges it to be the near-infrared marker, the visible light fill-in module is started, the acquiring module reacquires the face detection feature, and the setting module updates the detection feature mark and the detection feature source of the face detection feature.

[0052] A face recognition device includes a visible light camera, a near infrared camera, a central controller and a chassis with a hollow cavity.

[0053] The chassis includes a chassis body and a chassis independent chamber located on the top of the chassis body;

[0054] The visible light camera and near infrared camera are both placed in separate chambers of the chassis;

[0055] The central controller is arranged in the chassis body and is electrically connected to the visible light camera and the near infrared camera.

[0056] Preferably, the chassis also includes a chassis door panel movably connected to the chassis body.

[0057] Preferably, the independent chambers of the chassis are rotatable.

[0058] Furthermore, a visible light fill light and a near infrared fill light electrically connected to the central controller are provided in an independent chamber of the chassis.

[0059] Preferably, a shutter for exhaust and heat dissipation is provided at the rear of the chassis body.

[0060] Preferably, a voice player electrically connected to the central controller is provided on the chassis door panel.

[0061] The face recognition method provided by the present invention sets feature tags and feature sources for the registered face storage features and the detected face detection features respectively, that is, the feature tags and feature sources are used as identification bits. When performing specific face feature comparison, the identification bits can be compared first, avoiding direct comparison of face feature data with a larger number and volume, thereby speeding up the processing speed. Moreover, since the feature tags and feature sources are set separately, the light source is used as the main attribute of the feature tag and the device source is used as the main attribute of the feature source, giving the face features more distinguishing attributes, which can effectively avoid the interference of various abnormal face feature data. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0063] Figure 1 A schematic diagram of a flow chart of one implementation of a face recognition method in an embodiment of the present invention;

[0064] Figure 2 A structural diagram of one implementation of a face recognition system in an embodiment of the present invention;

[0065] Figure 3 The figure is a schematic diagram of the structure of one implementation mode of the face recognition device in the embodiment of the present invention. DETAILED DESCRIPTION

[0066] In order to enable those skilled in the art to better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments.

[0067] Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making any creative work should fall within the scope of protection of this application.

[0068] like Figure 1 As shown, a face recognition method includes the following steps:

[0069] S1. Register face storage features;

[0070] There are two ways to register facial storage features: the first way is that the administrator performs facial authentication and enters the user management menu, uses the acquisition device to collect the facial storage features of the current user, and the facial storage features are directly bound to the corresponding user, and the registration is completed. The second way is that the administrator uploads the user's photo, and the system collects the facial storage features in the user's photo, and directly binds it to the corresponding user, and the registration is completed.

[0071] These two methods can be used to directly register the facial storage features of the user himself, or to use photos to register facial storage features when the user himself cannot be present. At the same time, it can also support remote uploading of photos, which meets the requirements of binding the user with his facial storage features and completing the registration of facial storage features in various situations, regardless of whether the user himself is present or not, and whether the collection device is online or not, to form the facial data of this user.

[0072] There are many ways to bind facial storage features to users. In this embodiment, a user ID is set for each user, and the facial storage features of each user are matched with their respective user ID numbers to achieve a one-to-one binding effect. The specific steps will not be described in detail.

[0073] S2. Setting the storage feature flag and storage feature source of the face storage feature;

[0074] The storage feature mark includes a visible light mark and a near-infrared mark, that is, the storage feature mark itself is a mark bit, and the attributes of this mark bit include two attributes: a visible light mark and a near-infrared mark. That is, the storage feature mark of the same face storage feature can be only a visible light mark, only a near-infrared mark, or both a visible light mark and a near-infrared mark.

[0075] Similarly, the storage feature source of the same facial storage feature may be only the device source, only the photo source, or both the device source and the photo source.

[0076] S3. Obtain face detection features;

[0077] There are many ways to obtain the information, including obtaining it through video recording using a visible light camera and a near-infrared camera installed on the device, and obtaining it through photos using a visible light camera.

[0078] After capturing a face or photo, the specific method for obtaining face detection features may adopt the well-known technology in the art, such as collecting face image data, obtaining face vector features through calculation, and forming face detection features, which will not be described here.

[0079] S4. Setting the detection feature flag and detection feature source of the face detection feature;

[0080] According to the acquisition device and collection source of the face detection feature, the detection feature tag and the detection feature source are set. Just like setting the feature tag in S2, the detection feature tag of the same face detection feature can be only a visible light tag, only a near-infrared tag, or both a visible light tag and a near-infrared tag.

[0081] Similarly, the detection feature source of the same face detection feature may be only device source, only photo source, or both device source and photo source.

[0082] S5. Compare the storage feature source and the detection feature source, and compare the storage feature tag and the detection feature tag;

[0083] S6. If the comparison is successful, compare the stored face features and the face detection features;

[0084] The feature sources and feature tags are compared respectively. Since the amount of data contained in the feature sources and feature tags themselves is much smaller than that of the face storage features and face detection features, that is, the previous setting of the storage feature tags and storage feature sources divides all face storage features into several small categories. When comparing, the small categories are searched and compared first, and then the specific face storage features are compared. Therefore, compared with directly comparing the face storage features and face detection features, the workload of calculation or comparison is reduced, and the comparison efficiency is accelerated.

[0085] If the comparison is unsuccessful, for example, the acquisition equipment is operating normally or does not meet the environmental requirements, such as the dim cloudy day causing insufficient visible light intensity, making the face data collected by the visible light camera unclear, further resulting in invalid face detection feature acquisition, or insufficient clarity of the photo itself, which will also cause invalid face detection feature acquisition. In actual use, there will be various unexpected or special circumstances that cause invalid face detection feature acquisition or the stored feature tag and stored feature source to not meet the comparison rules. At this time, a prompt or warning message can be issued to remind the operator to check, and the entire authentication step will stop.

[0086] S7. If the facial storage features and facial detection features are successfully compared, the facial authentication is passed.

[0087] When the face storage features and face detection features meet the comparison requirements through calculation and comparison, for example, the similarity between the face detection features and the face storage features reaches the comparison threshold, the comparison is considered successful. At this time, face authentication is passed to complete the entire authentication process.

[0088] If the comparison between the stored facial features and the detected facial features fails, there are basically two situations. One is that the user is the same user, but because the user wears glasses, hats, etc. that will cause redundant detection features, the comparison threshold between the two cannot be reached. The other is that the detected user does not have the above situation, but the stored facial features of the user have not been registered before, or the corresponding data has been damaged after registration, which will also cause the comparison threshold between the two to fail.

[0089] Regardless of the reason why the comparison is unsuccessful, corresponding prompts or warning messages can be provided. The specific method and information content can be selected according to needs and will not be repeated here.

[0090] Preferably, step S2 specifically comprises:

[0091] S21. Set the storage feature source;

[0092] S22. If the storage feature source is a device source, the storage feature marker is set to a visible light marker and a near infrared marker;

[0093] S23. If the storage feature source is a photo source, the storage feature mark is set to a visible light mark.

[0094] If it is a photo source, it means that the facial storage features are registered through photo data. The near-infrared camera cannot directly obtain facial data from the photo itself. Only the visible light camera can obtain it directly from the photo. If it is a device source, it means that the facial storage features are registered through camera data collection. In this case, the near-infrared tag and the visible light tag are both matchable setting values.

[0095] Preferably, step S5 and step S6 are specifically:

[0096] S51. Compare the storage feature source and the detection feature source;

[0097] S52. If the comparison is successful, the storage feature mark and the detection feature mark are compared;

[0098] S61. If the storage feature mark and the detection feature mark are successfully compared, the face storage feature and the face detection feature are compared.

[0099] That is, step S5 can first compare the storage feature source and the detection feature source, and then compare the storage feature mark and the detection feature mark if the comparison is successful, and finally compare the face storage feature and the face detection feature. Compared with comparing the feature source and the feature mark at the same time, the comparison of the storage feature mark and the detection feature mark can be stopped after the comparison of the storage feature source and the detection feature source is unsuccessful, thereby saving comparison time.

[0100] Similarly, step S5 and step S6 are specifically as follows:

[0101] S53. Compare the stored signature and the detected signature;

[0102] S54. If the comparison is successful, the storage feature source and the detection feature source are compared;

[0103] S62. If the storage feature source and the detection feature source are successfully compared, the face storage features and the face detection features are compared.

[0104] Preferably, before step S5, the method further includes:

[0105] S41. Determine the detection feature mark;

[0106] S42. If it is a visible light mark, near-infrared fill light is performed to re-acquire the face detection feature, and the detection feature mark and detection feature source of the face detection feature are updated.

[0107] In some places with strict security or authentication requirements, it is necessary to improve the authentication strength without increasing the comparison threshold. In this case, visible light and near-infrared are used simultaneously. If there is only a visible light mark, it means that the previous work was done when the visible light intensity was sufficient. At this time, near-infrared fill light is used to re-acquire the face detection features based on the acquired near-infrared data and the previous visible light data, and update the detection feature mark and detection feature source.

[0108] By reacquiring near-infrared fill light, the types of face detection features involved in the comparison can be increased, the authentication strength can be improved, and relatively strict authentication requirements can be met.

[0109] Similarly, before step S5, the following steps are also included:

[0110] S43. Determine the source of the detection feature;

[0111] S44. If it is a near-infrared marker, perform visible light fill lighting, reacquire the face detection feature, and update the detection feature marker and detection feature source of the face detection feature.

[0112] If there is only a near-infrared mark, it means that the previous detection was performed when the visible light intensity was insufficient. At this time, visible light fill light is performed. Based on the acquired visible light data and the previous near-infrared data, the face detection features are re-obtained, and the detection feature mark and the detection feature source are updated at the same time. The effect is similar to the above and will not be repeated.

[0113] The present invention also provides a face recognition system, comprising:

[0114] Registration module 01, used to register face storage features;

[0115] The setting module 02 connected to the registration module 01 is used to set the storage feature mark and storage feature source of the face storage feature;

[0116] The acquisition module 03 connected to the setting module 02 is used to acquire the face detection features. The setting module 02 is also used to set the detection feature flag and the detection feature source of the face detection feature;

[0117] A comparison module 04 connected to the setting module 02, for comparing the storage feature source with the detection feature source, and for comparing the storage feature mark with the detection feature mark;

[0118] The comparison module 04 is also used to compare the face storage features and the face detection features;

[0119] The authentication module 05 connected to the comparison module 04 is used for face authentication.

[0120] In actual use, a terminal device such as a computer or a mobile phone can be used as an input device, and the existing processed facial storage features can be bound to the user one-to-one through the registration module 01.

[0121] The setting module 02, the comparison module and the authentication module 05 can all be installed on a chip or PLC with computing functions. It should also be noted that the connection in this embodiment does not limit the specific connection method, which can be a physical hard connection or a logical software connection, that is, it can be connected through a physical electrical connection or a wireless communication connection, such as through Bluetooth, infrared or wires, cables, etc., which will not be repeated here.

[0122] Preferably, it also includes:

[0123] A registration storage module connected between the registration module 01 and the setting module 02, the registration storage module is used to store face storage features;

[0124] The acquisition storage module is connected between the acquisition module 03 and the setting module 02, and is used to store face detection features.

[0125] The registration storage module and the acquisition storage module can select different storage devices according to actual needs. Since the amount of data that needs to be stored in the registration storage module is large, and the number of subsequent data supplements will be relatively small after a one-time batch initialization registration, a large-capacity traditional hard disk can be used. Of course, if the cost is not considered, a solid-state hard disk with faster reading speed but higher cost can also be used.

[0126] Since the storage module acquires a small amount of data each time, but the number of times the data is supplemented will be relatively large, or the frequency of updating the data will be relatively fast, a small-capacity solid-state hard drive is generally used. This will not cause excessive cost waste while meeting the reading speed and improving the effect efficiency.

[0127] Preferably, the acquisition module 03 includes a visible light camera component and a near infrared camera component, and generally adopts a visible light camera and a near infrared camera, that is, the acquisition module 03 adopts a dual camera mode of visible light and near infrared in order to adapt to more time and weather.

[0128] Furthermore, it also includes:

[0129] A determination module 06 connected to the acquisition module 03, for determining the detection characteristic mark;

[0130] A visible light fill light component connected to the visible light camera component;

[0131] A near-infrared fill light component connected to the near-infrared camera component;

[0132] If the judging module 06 judges that it is a visible light mark, the near infrared fill light component is activated, the acquiring module 03 reacquires the face detection feature, and the setting module 02 updates the detection feature mark and detection feature source of the face detection feature;

[0133] If the judging module 06 judges that it is a near-infrared marker, the visible light fill light component is started, the acquiring module 03 re-acquires the face detection feature, and the setting module 02 updates the detection feature mark and detection feature source of the face detection feature.

[0134] The visible light fill light component is a visible light fill light lamp, and the near-infrared fill light component is a near-infrared fill light lamp. The specific number of fill lights can be selected according to actual needs, and the adjustable fill light intensity can be set.

[0135] The present invention also discloses a face recognition device, comprising a visible light camera 1, a near infrared camera 2, a central controller 7 and a chassis with a hollow cavity.

[0136] The chassis includes a chassis body 6 and a chassis independent chamber located on the top of the chassis body 6;

[0137] The visible light camera 1 and the near infrared camera 2 are both arranged in independent chambers of the chassis;

[0138] The central controller 7 is disposed in the chassis body 6 and is electrically connected to the visible light camera 1 and the near infrared camera 2 .

[0139] This device adopts a visible light and near-infrared dual camera mode, that is, when registering facial storage features and obtaining facial detection features, the visible light camera 1 and near-infrared camera 2 set on the device are used to obtain them through video recording, and the visible light camera 1 is used to recognize photos to obtain them, so as to adapt to more demand environments.

[0140] The visible light camera 1 and the near-infrared camera 2 can be selected according to actual needs, and since they are both installed in independent chambers of the chassis, the visible light camera 1 and the near-infrared camera 2 can be quickly replaced or maintained separately without affecting other devices in the chassis body 6.

[0141] The central controller 7 is used to register the received facial storage features into the system, or store them in a memory such as a hard disk or a mobile hard disk, and to set the storage feature sources of the facial storage features accordingly, specifically the photo sources and the device sources, and to set the storage feature marks of the facial storage features accordingly, specifically the near-infrared marks and the visible light marks.

[0142] After processing the above steps, the feature sources and feature tags are respectively compared accordingly. Since the amount of data contained in the feature sources and feature tags themselves is much smaller than that of the face storage features and face detection features, that is, the previous setting of the storage feature tags and storage feature sources divides all face storage features into several small categories. When comparing, the small categories are searched and compared first, and then the specific face storage features are compared. Therefore, compared with directly comparing the face storage features and face detection features, the workload of calculation or comparison is reduced, and the comparison efficiency is accelerated.

[0143] If the comparison is unsuccessful, for example, the acquisition equipment is operating normally or does not meet the environmental requirements, such as the dim cloudy day causing insufficient visible light intensity, making the face data collected by the visible light camera 1 not clear enough, further resulting in invalid face detection feature acquisition, or insufficient clarity of the photo itself, which will also cause invalid face detection feature acquisition. In actual use, there will be various unexpected or special circumstances that cause invalid face detection feature acquisition or the stored feature tag and the stored feature source to not meet the comparison rules. At this time, a prompt or warning message can be issued to remind the operator to check, and the entire authentication step stops.

[0144] When the central controller 7 calculates and compares that the face storage features and the face detection features meet the comparison requirements, for example, the similarity between the face detection features and the face storage features reaches the comparison threshold, the comparison is considered successful. At this time, face authentication is passed to complete the entire authentication process.

[0145] In actual use, a liquid crystal display screen 5 can also be set on the chassis body 6 for administrators to perform human-computer interaction, or to display information on whether the authentication is successful or not, and can also be used to display information such as the current date and time, which will not be described in detail.

[0146] Preferably, the chassis also includes a chassis door panel movably connected to the chassis body 6.

[0147] The chassis door panel movably connected to the chassis body 6 can be opened when necessary to facilitate the maintenance or replacement of internal equipment. Different types of locks can be used to achieve the movable connection between the chassis body 6 and the chassis door panel according to different usage requirements. For example, the present embodiment uses a triangular lock for fixing.

[0148] Preferably, the independent chamber of the chassis is rotatable to adjust the angles of the visible light camera 1 and the near infrared camera 2, and can adapt to various installation environments to achieve good recognition effects. In addition, it also includes a chassis cover arranged on the upper part of the independent chamber of the chassis. When necessary, the visible light camera 1 and the near infrared camera 2 can be conveniently repaired or replaced by opening the chassis cover.

[0149] Furthermore, a visible light fill light 3 and a near infrared fill light 4 electrically connected to the central controller 7 are provided in the independent chamber of the chassis.

[0150] If the central processing unit determines that the detection feature mark is a visible light mark, the near-infrared fill light 4 is started to reacquire the face detection feature, and the detection feature mark and detection feature source of the face detection feature are updated;

[0151] If the detection feature mark is determined to be a near-infrared mark, the visible light fill light 3 is started to reacquire the face detection feature, and the detection feature mark and detection feature source of the face detection feature are updated.

[0152] The specific numbers of the visible light fill light 3 and the near-infrared fill light 4 can be selected according to actual needs, and the adjustable fill light intensity can be set.

[0153] The specific setting positions of the visible light fill light 3 and the near-infrared fill light 4 can be selected according to actual conditions. For example, in this embodiment, the visible light fill light 3 is on the left side close to the visible light camera 1, and the near-infrared fill light 4 is on the right side close to the near-infrared camera 2.

[0154] Preferably, a shutter for exhaust and heat dissipation is provided at the rear of the chassis body 6 to ensure good heat dissipation inside the chassis and ensure stable operation of the equipment.

[0155] Preferably, a voice player electrically connected to the central controller 7 is provided on the chassis door panel for voice notification of success or failure of acquisition. Of course, other types of information can also be played, such as time broadcast and weather broadcast.

[0156] The above is a detailed introduction to a face recognition method and system provided by the present invention. This article uses specific examples to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A face recognition method, characterized in that: The following steps are included: Register face storage features; Setting a storage feature tag and a storage feature source of the face storage feature; Get face detection features; Setting a detection feature tag and a detection feature source of the face detection feature; Comparing the stored feature source with the detected feature source, and comparing the stored feature tag with the detected feature tag; If the comparison is successful, then comparing the stored face features with the face detection features; If the facial storage features and the facial detection features are successfully compared, the facial authentication is passed; Among them, the storage feature mark and the detection feature mark include a visible light mark and a near infrared mark, and the storage feature source and the detection feature source include a device source and a photo source.

2. The face recognition method according to claim 1, characterized in that: The storage feature flag and storage feature source of setting the face storage feature are specifically: Setting the storage feature source; If the storage feature source is the device source, setting the storage feature marker to the visible light marker and the near infrared marker; If the storage feature source is the photo source, the storage feature mark is set to the visible light mark.

3. The face recognition method according to claim 1, characterized in that: The comparing the storage feature source and the detection feature source, and comparing the storage feature tag and the detection feature tag; if the comparison is successful, then comparing the face storage feature and the face detection feature is specifically: comparing the stored feature source and the detected feature source; If the comparison is successful, then comparing the stored characteristic mark and the detected characteristic mark; If the stored feature mark and the detected feature mark are successfully compared, then the stored face feature and the detected face feature are compared.

4. The face recognition method according to claim 1, characterized in that: The comparing the storage feature source and the detection feature source, and comparing the storage feature tag and the detection feature tag; if the comparison is successful, then comparing the face storage feature and the face detection feature is specifically: comparing the stored characteristic mark and the detected characteristic mark; If the comparison is successful, then comparing the stored feature source and the detected feature source; If the storage feature source and the detection feature source are successfully compared, the face storage feature and the face detection feature are compared.

5. The face recognition method according to claim 1, characterized in that: Before comparing the storage feature source with the detection feature source, and comparing the storage feature mark with the detection feature mark, the method further includes: Determining the detection feature mark; If it is a visible light mark, near-infrared fill light is performed to reacquire the face detection feature, and the detection feature mark and the detection feature source of the face detection feature are updated.

6. The face recognition method according to claim 1, characterized in that: Before comparing the storage feature source with the detection feature source, and comparing the storage feature mark with the detection feature mark, the method further includes: Determining the detection feature mark; If it is the near-infrared marker, visible light fill light is performed to reacquire the face detection feature, and the detection feature marker and the detection feature source of the face detection feature are updated.

7. A face recognition system, characterized in that: include: A registration module, used to register face storage features; A setting module connected to the registration module, used for setting the storage feature mark and storage feature source of the face storage feature; An acquisition module connected to the setting module and used to acquire face detection features, and the setting module is also used to set detection feature tags and detection feature sources of the face detection features; A comparison module connected to the setting module, used for comparing the storage feature source with the detection feature source, and for comparing the storage feature mark with the detection feature mark; The comparison module is also used to compare the face storage feature and the face detection feature; The authentication module connected to the comparison module is used to pass face authentication.

8. The face recognition system according to claim 7, characterized in that: Also includes: A registration storage module connected between the registration module and the setting module, the registration storage module is used to store the face storage features; An acquisition storage module connected between the acquisition module and the setting module, the acquisition storage module is used to store the face detection features.

9. The face recognition system according to claim 7, characterized in that: The acquisition module includes a visible light camera module and a near infrared camera module.

10. The face recognition system according to claim 9, characterized in that: Also includes: A determination module connected to the acquisition module, used for determining the source of the detection feature and the detection feature mark; A visible light fill light module connected to the visible light camera module; A near-infrared fill light module connected to the near-infrared camera module; If the judging module judges that it is a visible light mark, the near infrared fill light module is started, the acquiring module reacquires the face detection feature, and the setting module updates the detection feature mark and the detection feature source of the face detection feature; If the judging module judges it is a near-infrared mark, the visible light fill light module is started, the acquiring module reacquires the face detection feature, and the setting module updates the detection feature mark and the detection feature source of the face detection feature.

Images