Module for vehicle-mounted face recognition unlocking

Abstract

The invention discloses a module for vehicle-mounted face recognition unlocking, and relates to the field of vehicle-mounted face recognition unlocking, and the module comprises a mounting carrier, one side of the mounting carrier is fixedly connected with an outer stand column bottom plate, the other side of the mounting carrier is fixedly connected with an outer stand column cover plate, one side of the mounting carrier is fixedly provided with a separation ring, and the separation ring is fixedly connected with the outer stand column bottom plate. And a mounting connecting plate is fixedly mounted in the mounting carrier. The first camera and the second camera are both installed on the substrate, the electronic optical filters are arranged in the first camera and the second camera to filter light rays, and light with different wavelengths is selectively filtered, so that interference caused by a strong light source such as sunlight or strong light can be inhibited, the influence of an over-bright or over-dark area is reduced, and the brightness of the camera is improved. Light can be weakened when outdoor sunlight is strong, the problem of light channeling interference when face recognition is applied to an automobile is solved, and outdoor strong light resistance is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle-mounted face recognition unlocking, and in particular to a module for vehicle-mounted face recognition unlocking. Background Art

[0002] In-car face recognition unlocking is a technology used in car safety systems. It uses cameras and biometric algorithms to identify the driver's facial features to unlock the car.

[0003] For example, the "A vehicle-mounted face recognition device and method" with publication number CN115205927A includes: a face recognition unit; the face recognition unit is connected to an adjustment unit, the adjustment unit includes an adjustment mechanism, and the adjustment mechanism is fixedly connected to the outer end surface of the B-pillar of the car.

[0004] However, in the prior art, since the vehicle is in an outdoor environment and there are various ambient light sources around it, strong light interference is likely to occur during actual use. In addition, the fill light equipped with the camera comes into contact with outdoor light during the fill light process, which will further cause the light to become stronger, directly affecting the accuracy of facial image acquisition and easily causing light leakage problems, resulting in the facial recognition unlocking device being unable to normally recognize facial information in an outdoor environment. Summary of the invention

[0005] The purpose of the present invention is to provide a module for vehicle-mounted face recognition unlocking to solve the problem of light leakage raised in the above-mentioned background technology, which causes the face recognition unlocking device to be unable to normally recognize face information in outdoor environments.

[0006] To achieve the above-mentioned object, the present invention provides the following technical solutions: a module for vehicle-mounted face recognition unlocking, comprising a mounting carrier, one side of the mounting carrier is fixedly connected to an outer column bottom plate, the other side of the mounting carrier is fixedly connected to an outer column cover plate, one side of the mounting carrier is fixedly installed with a separator ring, the interior of the mounting carrier is fixedly installed with a mounting connection plate, a face recognition unlocking module is installed on the mounting connection plate, and the face recognition unlocking module comprises a camera unit, a computing unit and a storage unit; The camera unit includes a binocular module, a fill light module and a light shielding sleeve, wherein the binocular module is fixedly mounted on one side of the mounting connection plate, the fill light module is fixedly mounted on the top of the binocular module, and the light shielding sleeve is fixedly mounted on the edge of the binocular module; The binocular module includes a substrate, a first camera and a second camera, wherein the first camera and the second camera are symmetrically mounted at two ends of the substrate, and one end of the first camera and one end of the second camera are both fixedly connected with an electronic filter; The fill light module includes a carrying plate, a fill light and a light intensity sensor. The fill light is fixedly installed at the center of the carrying plate. The fill light is located on the symmetry line between the first camera and the second camera. The light intensity sensor is located on one side of the fill light.

[0007] Preferably, the computing unit includes an embedded computing device, a face recognition algorithm, and unlocking and rejecting; The embedded computing device is used to process image data and split facial information into multiple fragments; The face recognition algorithm is used to perform face detection and feature extraction; The unlocking and rejecting are used to determine the opening and closing of the vehicle door.

[0008] Preferably, the face recognition algorithm includes a user interface unit, an image acquisition unit, a face detection unit, a feature extraction unit and a feature comparison unit; The user interface unit is used for users to input information and log in users; The image acquisition unit is used to acquire facial images; The face detection unit is used to perform data detection on the collected face image; The feature extraction unit is used to extract features in the data that are highly recognizable and difficult to forge; The feature comparison unit is used to compare the feature with the data in the database to determine the similarity of the features.

[0009] Preferably, the user interface includes a user registration unit, a face recognition unlocking unit and a result prompting unit; When the user registration unit is used for user registration, the system collects the user's facial image data, pre-processes the image, extracts the facial feature data and stores it in the database; The face recognition unlocking unit is used for the system to obtain user images in real time, perform face detection, capture faces, extract face features in the image, and compare them with user features stored in the database; The result prompting unit is used to prompt the recognition result. If the recognition is successful, the unlocking operation is performed. If the recognition fails, "recognition failed" is prompted and the lock is maintained.

[0010] Preferably, the step of extracting facial feature data and storing it in a database further comprises the following steps: S1. Data encryption: Encryption is performed when storing user characteristics to prevent data leakage; S2, anti-counterfeiting mechanism, using live mechanism to avoid photo deception; S3, data optimization, recording facial data under different lighting, angles and facial expressions.

[0011] Preferably, in step S1, the data encryption further comprises the following steps: S11, security test, input photos, videos and masks related to the user for verification to simulate different attack methods; S12. Update algorithms regularly, use CNNs, FaceNet, ArcFace, and SphereFace to calculate face data, and upgrade algorithms regularly.

[0012] Preferably, in step S2, the anti-counterfeiting mechanism further comprises the following steps: S21, liveness detection, putting the input face image into Euclidean space and analyzing its posture, position and motion trajectory information; S22. Multimodal recognition combines iris features, pupil features and lip features for multiple verifications.

[0013] Preferably, in step S21, the liveness detection further comprises the following steps: S211, dynamic analysis of facial features, detecting dynamic features of the user's blinking, micro and head movements; S212, 3D depth perception, using the camera to capture the depth information of the face; S213, infrared light, uses infrared light to capture changes in blood flow and skin temperature; S214: Execute specific actions, such as turning the head and blinking within a certain period of time.

[0014] Preferably, S31, generating an adversarial network, generating high-quality synthetic data, and simultaneously enhancing the diversity of the training set to improve the recognition accuracy of the model; S32, transfer learning, using deep learning models to fine-tune face recognition and perform efficient learning on small data sets.

[0015] Preferably, the storage unit is used to store the data in a database for comparison with the detection data at any time, and the storage unit is signal-connected to the calculation unit.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. In the present invention, the first camera and the second camera are both mounted on a substrate, and an electronic filter is arranged inside the substrate to filter the light. By selectively filtering light of different wavelengths, the interference caused by strong light sources such as sunlight or strong lamps can be suppressed, and the influence of overly bright or overly dark areas can be reduced. The light can be weakened when the sunlight is strong outdoors, and the electronic filter can collect multiple images with different lighting conditions. By combining the high-quality parts of these images, a high-quality final image is obtained to prevent the collection of facial images from being interfered by strong light, solve the problem of light interference in facial recognition applications in automobiles, and significantly improve the outdoor anti-strong light performance. The fill light is always located on the symmetry line between the first camera and the second camera. The fill light in the fill light module is used to supplement the light, so that a clear face can be obtained even in the dark. The light intensity sensor arranged on the side is used to detect the light intensity in real time to ensure that the fill light does not emit too strong light to affect the facial image.

[0017] 2. In the present invention, the embedded computing device can improve the overall performance of the system. The face recognition algorithm can process and identify a large amount of image data in a short time to ensure fast unlocking and response. The face recognition algorithm can adapt to different environments and user changes, including changes in lighting, age growth, etc., to enhance the adaptability of the system. Unlocking and rejection can provide multi-level security protection, and successfully identified and unidentified users are processed separately to ensure that only authorized users can unlock the device.

[0018] 3. In the present invention, data encryption is used to protect the user's facial data from unauthorized access, and the user's privacy is maintained by encrypting the facial data. Liveness detection technology is set to distinguish real users from fake materials such as photos, videos or masks, thereby improving recognition accuracy and security. Through fast and effective liveness detection, users can complete identity authentication with simple operations such as blinking and smiling. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic diagram of the three-dimensional structure of a mounting carrier used in a module for vehicle-mounted face recognition unlocking according to the present invention; Figure 2 This is a schematic diagram of the planar structure of a binocular module and a fill light module used in a module for vehicle-mounted face recognition unlocking of the present invention; Figure 3 A schematic diagram of the internal structure of a mounting carrier used in a module for vehicle-mounted face recognition unlocking according to the present invention; Figure 4 A block diagram of a face recognition module for vehicle-mounted face recognition unlocking according to the present invention; Figure 5 A flow chart of a face recognition algorithm of a module for vehicle-mounted face recognition unlocking according to the present invention; Figure 6 A block diagram of a user interface of a module for vehicle-mounted face recognition unlocking according to the present invention; Figure 7 The present invention is a block diagram of a module for vehicle-mounted face recognition unlocking, which extracts face feature data and stores it in a database.

[0020] In the figure: 1. Installation carrier; 2. Separation ring; 3. Binocular module; 4. Fill light module; 5. Shading sleeve; 6. Installation connecting plate; 7. External column base plate; 8. External column cover plate; 9. Computing unit; 10. Storage unit; 11. Embedded computing device; 12. Face recognition algorithm; 13. Unlock and reject; 14. User interface unit; 15. Image acquisition unit; 16. Face detection unit; 17. Feature extraction unit; 18. Feature comparison unit; 19. User registration unit; 20. Face recognition unlocking unit; 21. Result prompt unit; 31. Base plate; 32. First camera; 33. Second camera; 34. Electronic filter; 41. Carrier plate; 42. Fill light; 43. Light intensity sensor. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the implementation regulations described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] Example 1: Reference Figure 1 , Figure 2 and Figure 3 As shown: a module for vehicle-mounted face recognition unlocking, comprising a mounting carrier 1, one side of the mounting carrier 1 is fixedly connected to an outer column bottom plate 7, the other side of the mounting carrier 1 is fixedly connected to an outer column cover plate 8, one side of the mounting carrier 1 is fixedly installed with a separator ring 2, the interior of the mounting carrier 1 is fixedly installed with a mounting connecting plate 6, the mounting connecting plate 6 is installed with a face recognition unlocking module, and the face recognition unlocking module comprises a camera unit, a computing unit 9 and a storage unit 10; The camera unit includes a binocular module 3, a fill light module 4 and a light shielding sleeve 5. The binocular module 3 is fixedly mounted on the surface of the mounting carrier 1, the fill light module 4 is fixedly mounted on the top of the binocular module 3, and the light shielding sleeve 5 is fixedly mounted on the edge of the binocular module 3. The binocular module 3 includes a substrate 31, a first camera 32 and a second camera 33. The first camera 32 and the second camera 33 are symmetrically mounted at two ends of the substrate 31. One end of the first camera 32 and one end of the second camera 33 are both fixedly connected with an electronic filter 34. The fill light module 4 includes a carrier plate 41 , a fill light 42 and a light intensity sensor 43 . The fill light 42 is fixedly mounted at the center of the carrier plate 41 . The fill light 42 is located on the symmetry line between the first camera 32 and the second camera 33 . The light intensity sensor 43 is located on one side of the fill light 42 .

[0023] In this embodiment, the mounting carrier 1 is fixedly mounted on the vehicle body, and its outer side is protected by the outer column cover plate 8, and its inner side is protected by the outer column bottom plate 7, so as to ensure that the mounting carrier 1 will not be violently damaged on the vehicle body, and the separation ring 2 is used to separate an area for installing the face recognition module; The face recognition unlocking module includes a camera unit, a computing unit 9 and a storage unit 10, wherein the camera unit is used to obtain face data, the computing unit 9 is used to analyze and determine the similarity of face data, and the storage unit 10 is used to store the data in a database; The binocular module 3 in the camera unit uses the first camera 32 and the second camera 33 to identify the user's facial information, and can also be used to obtain eye information simultaneously. The first camera 32 and the second camera 33 are both installed on the substrate 31, and an electronic filter 34 is arranged inside to filter the light. By selectively filtering light of different wavelengths, the interference caused by strong light sources such as sunlight or strong lights can be suppressed, and the influence of overly bright or dark areas can be reduced. The light can be weakened when the sunlight is strong outdoors, and the electronic filter 34 can collect multiple images with different lighting conditions. By combining the high-quality parts of these images, a high-quality final image is obtained to prevent the facial image collection from being interfered by strong light, solve the problem of light interference in the application of face recognition in automobiles, and significantly improve the outdoor anti-strong light performance; A fill light module 4 is arranged on the top of the binocular module 3. A carrying plate 41 in the fill light module 4 is connected to the base plate 31 to ensure that the fill light 42 is always located on the symmetry line between the first camera 32 and the second camera 33. The fill light 42 in the fill light module 4 is used to supplement light, so that a clear face can be obtained even in a dark place. A light intensity sensor 43 arranged on the side thereof is used to detect the light intensity in real time to ensure that the fill light 42 does not emit too strong light and affect the face image.

[0024] Embodiment 2: According to Figure 5 As shown, the computing unit 9 includes an embedded computing device 11, a face recognition algorithm 12, and an unlock and reject algorithm 13; The embedded computing device 11 is used to process the image data and split the facial information into multiple fragments; The face recognition algorithm 12 is used to perform face detection and feature extraction; Unlock and Reject 13 are used to confirm the opening and closing of the door; The face recognition algorithm 12 includes a user interface unit 14, an image acquisition unit 15, a face detection unit 16, a feature extraction unit 17 and a feature comparison unit 18; The user interface unit 14 is used for users to input information and log in users; The image acquisition unit 15 is used to acquire facial images; The face detection unit 16 is used to perform data detection on the collected face image; The feature extraction unit 17 is used to extract features in the data that are highly recognizable and difficult to forge; The feature comparison unit 18 is used to compare the feature with the data in the database to determine the similarity of the features.

[0025] In this embodiment, the embedded computing device 11 can respond quickly, reduce delays, and improve the overall performance of the system. The face recognition algorithm 12 can process and identify a large amount of image data in a short time to ensure fast unlocking and response, and the face recognition algorithm 12 can be continuously updated and learned to adapt to different environments and user changes, including changes in lighting, age growth, etc., to enhance the adaptability of the system. Unlocking and rejection 13 can provide multi-level security protection, and handle successfully identified and unidentified users separately to ensure that only authorized users can unlock the device while preventing unauthorized access. Through real-time image processing, it can quickly respond to user operations and optimize the overall interactive experience.

[0026] Embodiment 3: According to Figure 6 and Figure 7 As shown, the user interface unit 14 includes a user registration unit 19, a face recognition unlocking unit 20 and a result prompting unit 21; The user registration unit 19 is used for user registration. When the system collects the user's facial image data, pre-processes the image, extracts the facial feature data and stores it in the database; The face recognition unlocking unit 20 is used for the system to obtain user images in real time, perform face detection, capture faces, extract face features in the image, and compare them with user features stored in the database; The result prompting unit 21 is used to prompt the recognition result. If the recognition is successful, the unlocking operation is performed. If the recognition fails, the "recognition failed" prompt is given and the lock is maintained; Extracting facial feature data and storing it in a database also includes the following steps: Step 1: Data encryption: Encrypt when storing user characteristics to prevent data leakage; Step 2: Anti-counterfeiting mechanism, using live mechanism to avoid photo deception; Step 3: Data optimization, recording facial data under different lighting, angles and facial expressions; In step one, data encryption also includes the following steps: Security testing: input user-related photos, videos, and masks for verification to simulate different attack methods; Update algorithms regularly, use CNNs, FaceNet, ArcFace, and SphereFace to calculate face data, and upgrade algorithms regularly.

[0027] In step 2, the anti-counterfeiting mechanism also includes the following steps: Liveness detection: put the input face image into Euclidean space and analyze its posture, position and motion trajectory information; Multimodal recognition, combining iris features, pupil features, and lip features for multiple verifications; Liveness detection also includes the following steps: Dynamic facial feature analysis to detect the user's blinking, micro-movements and head movements; 3D depth perception, using cameras to capture depth information of the face; Infrared lighting, which uses infrared light to capture changes in blood flow and skin temperature; Perform specific actions, such as turning your head and blinking your eyes within a certain time; In step three, data optimization also includes the following steps: Generate adversarial networks to generate high-quality synthetic data and simultaneously enhance the diversity of training sets to improve the recognition accuracy of the model; Transfer learning, using deep learning models to fine-tune face recognition and learn efficiently on small datasets In this embodiment, data encryption protects the user's facial data from unauthorized access. By encrypting the facial data, user privacy is maintained. Liveness detection technology can effectively distinguish between real users and fake materials such as photos, videos or masks, thereby improving recognition accuracy and security. Through fast and effective liveness detection, users can complete identity authentication with simple operations such as blinking and smiling, thereby improving user experience. Data optimization technology can improve the efficiency and accuracy of face recognition algorithm 12 by removing redundant and unnecessary data. The optimized data processing speed is faster, which significantly reduces the response time of face recognition in actual applications, helping the system to better adapt to different environments and user conditions such as changes in lighting, angles, etc., so that the recognition performance remains stable in a variety of situations.

[0028] The present invention uses the first camera 32 and the second camera 33 to identify the user's facial information, and can also be used to obtain eye information simultaneously. The first camera 32 and the second camera 33 are internally provided with an electronic filter 34 to filter the light. By selectively filtering light of different wavelengths, the interference caused by strong light sources such as sunlight or strong lights can be suppressed, and the influence of overly bright or overly dark areas can be reduced. When the outdoor sunlight is strong, the light can be weakened, and multiple images with different lighting conditions can be collected. By combining the high-quality parts of these images, a high-quality final image is obtained to prevent the facial image collection from being interfered by strong light, solve the problem of light interference in the application of facial recognition in automobiles, and significantly improve the outdoor anti-strong light performance; The fill light 42 is always located on the symmetry line between the first camera 32 and the second camera 33. The fill light 42 in the fill light module 4 is used to supplement light so as to obtain a clear face even in a dark place. The light intensity sensor 43 arranged on the side thereof is used to detect the light intensity in real time to ensure that the fill light 42 does not emit too strong light to affect the face image. The embedded computing device 11 of the present invention can respond quickly, reduce delays, and improve the overall performance of the system. The face recognition algorithm 12 can process and identify a large amount of image data in a short time to ensure fast unlocking and response, and the face recognition algorithm 12 can be continuously updated and learned to adapt to different environments and user changes, including changes in lighting, age growth, etc., to enhance the adaptability of the system. Unlocking and rejection 13 can provide multi-level security protection, and process successfully identified and unidentified users separately to ensure that only authorized users can unlock the device while preventing unauthorized access. Through real-time image processing, it can quickly respond to user operations and optimize the overall interactive experience. Data encryption protects the user's facial data from unauthorized access. By encrypting the facial data, user privacy is maintained. Liveness detection technology can effectively distinguish real users from forged materials such as photos, videos or masks, and improve recognition accuracy and security. Through fast and effective liveness detection, users can complete identity authentication with simple operations such as blinking and smiling.

[0029] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A module for vehicle-mounted face recognition unlocking, comprising a mounting carrier (1), one side of the mounting carrier (1) being fixedly connected to an outer column bottom plate (7), the other side of the mounting carrier (1) being fixedly connected to an outer column cover plate (8), and one side of the mounting carrier (1) being fixedly mounted to a spacer ring (2), characterized in that: A mounting connection plate (6) is fixedly mounted inside the mounting carrier (1), a face recognition unlocking module is mounted on the mounting connection plate (6), and the face recognition unlocking module comprises a camera unit, a computing unit (9) and a storage unit (10); The camera unit comprises a binocular module (3), a fill light module (4) and a light shielding sleeve (5); the binocular module (3) is fixedly mounted on one side of a mounting connection plate (6); the fill light module (4) is fixedly mounted on the top of the binocular module (3); and the light shielding sleeve (5) is fixedly mounted on the edge of the binocular module (3); The binocular module (3) comprises a substrate (31), a first camera (32) and a second camera (33), wherein the first camera (32) and the second camera (33) are symmetrically mounted at two ends of the substrate (31), and one end of the first camera (32) and one end of the second camera (33) are both fixedly connected to an electronic filter (34); The fill light module (4) comprises a carrier plate (41), a fill light (42) and a light intensity sensor (43); the fill light (42) is fixedly mounted at the center of the carrier plate (41); the fill light (42) is located on a symmetry line between the first camera (32) and the second camera (33); and the light intensity sensor (43) is located on one side of the fill light (42).

2. The module for vehicle-mounted face recognition unlocking according to claim 1, characterized in that: The computing unit (9) includes an embedded computing device (11), a face recognition algorithm (12), and an unlock and reject function (13); The embedded computing device (11) is used to process image data and split facial information into multiple fragments; The face recognition algorithm (12) is used to perform face detection and feature extraction; The unlocking and rejecting (13) are used to determine the opening and closing of the vehicle door.

3. The module for vehicle-mounted face recognition unlocking according to claim 2, characterized in that: The face recognition algorithm (12) comprises a user interface unit (14), an image acquisition unit (15), a face detection unit (16), a feature extraction unit (17) and a feature comparison unit (18); The user interface unit (14) is used for users to input information and log in as users; The image acquisition unit (15) is used to acquire facial images; The face detection unit (16) is used to perform data detection on the collected face image; The feature extraction unit (17) is used to extract features in the data that are highly recognizable and difficult to forge; The feature comparison unit (18) is used to compare the feature with the data in the database to determine the degree of similarity of the features.

4. The module for vehicle-mounted face recognition unlocking according to claim 3, characterized in that: The user interface unit (14) comprises a user registration unit (19), a face recognition unlocking unit (20) and a result prompting unit (21); The user registration unit (19) is used for user registration. When the system collects the user's facial image data, pre-processes the image, extracts the facial feature data and stores it in a database; The face recognition unlocking unit (20) is used for the system to obtain a user image in real time, perform face detection, capture the face, extract the face features in the image, and compare them with the user features stored in the database; The result prompting unit (21) is used to prompt the recognition result. If the recognition is successful, an unlocking operation is performed. If the recognition fails, "recognition failed" is prompted and the lock is maintained.

5. The module for vehicle-mounted face recognition unlocking according to claim 4, characterized in that: The extracting of facial feature data and storing it in a database further comprises the following steps: S1. Data encryption: Encryption is performed when storing user characteristics to prevent data leakage; S2, anti-counterfeiting mechanism, using live mechanism to avoid photo deception; S3, data optimization, recording facial data under different lighting, angles and facial expressions.

6. The module for vehicle-mounted face recognition unlocking according to claim 5, characterized in that: In step S1, the data encryption further includes the following steps: S11, security test, input photos, videos and masks related to the user for verification to simulate different attack methods; S12. Update algorithms regularly, use CNNs, FaceNet, ArcFace, and SphereFace to calculate face data, and upgrade algorithms regularly.

7. The module for vehicle-mounted face recognition unlocking according to claim 5, characterized in that: In step S2, the anti-counterfeiting mechanism further comprises the following steps: S21, liveness detection, putting the input face image into Euclidean space and analyzing its posture, position and motion trajectory information; S22. Multimodal recognition combines iris features, pupil features and lip features for multiple verifications.

8. The module for vehicle-mounted face recognition unlocking according to claim 7, characterized in that: In step S21, the liveness detection further includes the following steps: S211, dynamic analysis of facial features, detecting dynamic features of the user's blinking, micro and head movements; S212, 3D depth perception, using the camera to capture the depth information of the face; S213, infrared light, uses infrared light to capture changes in blood flow and skin temperature; S214: Execute specific actions, such as turning the head and blinking within a certain period of time.

9. The module for vehicle-mounted face recognition unlocking according to claim 5, characterized in that: In step S3, the data optimization further includes the following steps: S31, Generate adversarial networks to generate high-quality synthetic data, and simultaneously enhance the diversity of training sets to improve the recognition accuracy of the model; S32, transfer learning, using deep learning models to fine-tune face recognition and perform efficient learning on small data sets.

10. The module for vehicle-mounted face recognition unlocking according to claim 1, characterized in that: The storage unit (10) is used to store data in a database for comparison with detection data at any time, and the storage unit (10) is signal-connected to the calculation unit (9).

Images