Driving face image encryption identification system and method

Abstract

The invention provides a driving face image encryption identification system and method. The driving face image encryption recognition system comprises a display and an encrypted image recognition server. The display includes an optical encryption element and an image extraction device. The optical encryption element is provided with a plurality of light holes and is used for receiving reflected light reflected by the face of a driver and forming encrypted light through the light holes. The image extraction device is used for receiving the encrypted light and forming an encrypted image. The encrypted image recognition server is used for receiving an encrypted image and comprises an encrypted image feature extraction module and an encrypted image recognition module. The encrypted image feature extraction module is used for extracting encryption features of the encrypted image. The encrypted image identification module is used for comparing the encrypted feature with the registered encrypted feature to judge whether the encrypted feature is the registered encrypted feature of the registered driving, and generating identification result data according to the encrypted feature and the registered encrypted feature.

Description

Technical Field

[0001] The present disclosure relates to a driving face image encryption and identification system and method, and particularly to a driving face image encryption and identification system and method for encrypting a face image through an optical encryption element. Background Art

[0002] In current driving identification systems, identification systems with lenses are mostly used. However, the large volume of the lenses causes problems when the driving identification system is combined with a display, as the volume cannot be reduced. This is particularly prominent in the case of limited vehicle space, where the large volume of the display causes space constraints. In addition, many face identification methods verify by obtaining the face image of the driver, which may lead to the risk of the driver's face image leaking due to security vulnerabilities. Therefore, the large volume occupied by the lenses and privacy concerns are problems that need to be solved in current driving identification systems. Summary of the Invention

[0003] The present disclosure provides a driving face image encryption and identification system, including a display and an encrypted image identification server. The display is arranged in the direction facing a driver's face and includes an optical encryption element and an image extraction device. The optical encryption element has a plurality of light-transmitting holes and is used to receive the reflected light reflected by the driver's face and form encrypted light through the plurality of light-transmitting holes. The image extraction device is used to receive the encrypted light and form an encrypted image. The encrypted image identification server is communicatively connected to the image extraction device and is used to receive the encrypted image. The encrypted image identification server includes an encrypted image feature extraction module and an encrypted image identification module. The encrypted image feature extraction module is used to extract the encrypted features of the encrypted image. The encrypted image identification module is used to compare the encrypted features with the registered encrypted features to determine whether the encrypted features are the registered encrypted features of a registered driver, and generate identification result data accordingly.

[0004] In one embodiment, the encrypted image identification server further includes a vehicle locking module, which is used to unlock the vehicle function when the identification result data shows that the encrypted features are the registered encrypted features.

[0005] In one embodiment, the driving face image encryption and identification system further includes a registered driver database, which is communicatively connected to the encrypted image identification server and is used to store the encrypted features corresponding to the face extracted by the encrypted image feature extraction module when the driver is registered as a registered driver as the registered encrypted features. The encrypted image identification module accesses the registered driver database to compare the encrypted features with the registered encrypted features.

[0006] In one embodiment, the display further includes a transmissive display panel, which is disposed between the optical encryption element and the face of the driver, so that the reflected light passes through and is then received by the optical encryption element. The transmissive display panel is communicatively connected to the encrypted image recognition server and is used to receive the recognition result data and display a corresponding recognition result.

[0007] In one embodiment, the display further includes an optical element, which is used to emit light to the face of the driver. The light is visible light or infrared light, and the light reflects to produce reflected light after irradiating the face.

[0008] In one embodiment, the first distance between the face of the driver and the optical encryption element is at least 150 times the second distance between the optical encryption element and the image extraction device.

[0009] In one embodiment, the encrypted image recognition module compares the encrypted feature with the registered encrypted feature, calculates the similarity according to the distance between the feature vectors of the encrypted feature and the registered encrypted feature, and determines whether the encrypted feature is a registered encrypted feature according to whether the similarity is higher than a preset value.

[0010] The present disclosure further provides a method for encrypting and recognizing a driver's face image, including: receiving the reflected light reflected by the driver's face so that the reflected light forms encrypted light through a plurality of light-transmitting holes of the optical encryption element; forming an encrypted image from the encrypted light; extracting the encrypted feature of the encrypted image; and comparing the encrypted feature with the registered encrypted feature to determine whether the encrypted feature is the registered encrypted feature of a registered driver, and generating recognition result data accordingly.

[0011] In one embodiment, the registered encrypted feature is the encrypted feature corresponding to the face extracted when the driver is registered as a registered driver.

[0012] In one embodiment, the method for encrypting and recognizing a driver's face image further includes: unlocking a vehicle function when the recognition result data indicates that the encrypted feature is a registered encrypted feature. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] To more fully understand the embodiments and their advantages, reference is now made to the following description taken in conjunction with the accompanying drawings, in which:

[0014] Figure 1 is a schematic diagram of a system for encrypting and recognizing a driver's face image according to an embodiment of the present disclosure;

[0015] Figure 2 is a schematic diagram of the driver registration process in the method for encrypting and recognizing a driver's face image according to an embodiment of the present disclosure; and

[0016] Figure 3 is a schematic diagram of the driver recognition process in the method for encrypting and recognizing a driver's face image according to an embodiment of the present disclosure.

[0017] The reference numerals are as follows:

[0018] 100: Driving face image encryption and identification system

[0019] 110: Display

[0020] 111: Light-transmitting plate

[0021] 112: Light-transmitting display panel

[0022] 113, 113F: Optical encryption element

[0023] 114: Image extraction device

[0024] 120: Encrypted image identification server

[0025] 121: Feature extraction module

[0026] 122: Encrypted image identification module

[0027] 123: Vehicle locking module

[0028] 130: Registered driver database

[0029] 200: Driver registration process

[0030] 300: Driver identification process

[0031] D1: First distance

[0032] D2: Second distance

[0033] EI: Encrypted image

[0034] F: Face

[0035] FI: Face image

[0036] H: Light-transmitting hole

[0037] L1: Reflected light

[0038] L2: Encryption light

[0039] S201~S204, S301~S307: Steps Detailed implementation manners

[0040] Embodiments of the present disclosure are discussed in detail below. However, it can be understood that the embodiments provide many applicable concepts that can be implemented in various specific contexts. The discussed and disclosed embodiments are for illustration only and are not intended to limit the scope of the present disclosure.

[0041] Figure 1 It is a schematic diagram of a driving face image encryption and identification system 100 according to an embodiment of the present disclosure. AsFigure 1 As shown, the driving face image encryption and recognition system 100 includes a display 110, an encrypted image recognition server 120, and a registered driver database 130. The display 110 includes a light-transmitting plate 111, a light-transmitting display panel 112, an optical encryption element 113, and an image extraction device 114. The encrypted image recognition server 120 includes a feature extraction module 121, an encrypted image recognition module 122, and a vehicle locking module 123. The encrypted image recognition server 120 is communicatively connected to the light-transmitting display panel 112, the image extraction device 114, and the registered driver database 130. The display 110 is configured to be arranged in the order of the light-transmitting plate 111, the light-transmitting display panel 112, the optical encryption element 113, and the image extraction device 114 in front of the face F of the driver. The face F is at a first distance D1 from the optical encryption element 113, and the optical encryption element 113 is at a second distance D2 from the image extraction device 114.

[0042] Specifically, the display 110 is a vehicle-mounted display, and the light-transmitting display panel 112 is used to display vehicle-related images, such as vehicle status, audio-visual content, or the recognition result of the encrypted image recognition module 122 for the face F of the driver. The light-transmitting plate 111 is used to protect the light-transmitting display panel 112.

[0043] One plane of the optical encryption element 113 facing the driver is as Figure 1 shown by the optical encryption element 113F of. There are multiple light-transmitting holes H on the optical encryption element 113F, and both the light-transmitting plate 111 and the light-transmitting display panel 112 are light-transmitting. When the reflected light L1 of the face F of the driver passes through the light-transmitting plate 111 and the light-transmitting display panel 112, it will irradiate the optical encryption element 113, and the optical encryption element 113 receives it and forms an encrypted light L2 through the light-transmitting holes H. In an embodiment, the display 110 can also be equipped with an optical element that can emit visible light or infrared light to irradiate the face F of the driver and generate the reflected light L1. In the case without an optical element, the ambient light is used to generate the reflected light L1 to pass through the optical encryption element 113.

[0044] The optical encryption element 113 is a two-dimensional pinhole array. The encryption method of the encrypted image EI can be adjusted by adjusting the distribution of the light-transmitting holes H of the optical encryption element 113, such as the size and position of the light-transmitting holes H, to obtain encrypted images EI of different images. It can also be adjusted according to the resolution of the encrypted image EI, or the distribution of the light-transmitting holes H corresponding to the highest recognition accuracy is taken as the distribution of the light-transmitting holes H of the optical encryption element 113 used in this situation according to the recognition result of the encrypted image recognition module 122.

[0045] The image extraction device 114 is located behind the optical encryption element 113 to receive the encrypted light beam L2 to form an encrypted image EI, and then transmits it to the encrypted image identification server 120 for driver registration or identification. Therefore, as Figure 1 shown, an image not encrypted by the optical encryption element 113, such as the face image FI, will display a normal face picture, while after passing through the optical encryption element 113, the original face F picture can no longer be seen, so as to achieve the effect of image encryption. Accordingly, using the optical encryption element 113 to replace the original lens can reduce the occupied volume in the display 110 compared with the lens, and can also encrypt the image. In addition, the encrypted light beam L2 can also use other optical elements to change the light path to reach the image extraction device 114. The present disclosure does not limit the image extraction device 114 to be directly behind the optical encryption element 113.

[0046] The relationship between the first distance D1 of the driver's face F from the optical encryption element 113 and the second distance D2 between the optical encryption element 113 and the image extraction device 114. In one embodiment, the first distance D1 is set to be much greater than the second distance D2. For example, the first distance D1 can be 150 to 200 times the second distance D2. As long as the encrypted image EI for driver identification can be obtained, the present disclosure does not particularly limit the first distance D1 and the second distance D2.

[0047] The encrypted image identification server 120 is used to perform encrypted feature extraction on the received encrypted image EI by the feature extraction module 121. In the case of driver registration, the extracted encrypted features will be transmitted to the registered driver database 130 for storage. In the case of using it to identify whether the driver is a registered driver, the encrypted image identification module 122 determines whether there are registered encrypted features in the registered driver database 130 that are the same as the encrypted features, that is, compares the encrypted features with the registered encrypted features in the registered driver database 130 to determine whether the driver has the permission to unlock the vehicle. In one embodiment, an identification result data can be generated according to the identification result and transmitted to the vehicle locking module 123. In order to enable the driver to immediately know the identification result, the encrypted image identification module 122 can also transmit the identification result data to the transmissive display panel 112 for display.

[0048] The feature extraction module 121 extracts encrypted features of the encrypted image EI using a feature extraction model, and the encrypted image identification module 122 determines whether the encrypted features correspond to registered encrypted features using a classification model. The feature extraction module 121 and the encrypted image identification module 122 can use two different machine learning models. After one model extracts encrypted features from the encrypted image EI, another model can identify whether the corresponding driver is a registered driver based on the encrypted features. Alternatively, the same machine learning model can simultaneously extract encrypted features from the encrypted image EI and identify whether the corresponding driver is a registered driver. The model can be trained by inputting a large amount of face image FI data and training the model multiple times in a scenario simulation manner to identify whether these face images FI correspond to the face F of a registered driver.

[0049] In one embodiment, the identification method can calculate the similarity based on the distance between the encrypted features and the feature vectors of the registered encrypted features, and consider two encrypted features with a similarity higher than a preset value to correspond to the same face image FI. The machine learning model can be, for example, a convolutional neural network (CNN), ResNet50, etc. Different machine learning models can be selected according to the computing cost and identification accuracy, as long as they can be used to extract encrypted features from images and identify whether the driver is a registered driver. The present disclosure does not particularly limit this.

[0050] Accordingly, after the face image FI is encrypted by the optical encryption element 113 and the feature extraction module 121 extracts the features, it is difficult to know the actual picture of the face image FI from the encrypted features or the registered encrypted features. Whether the registered encrypted features are stored in the registered driver database 130 or identification comparison is performed, privacy protection is enhanced.

[0051] The vehicle locking module 123 receives the identification result data from the encrypted image identification module 122 and determines whether to unlock the vehicle function according to the result shown in the identification result data. If the identification result data shows that this driver is a registered driver, the vehicle function is unlocked; otherwise, the locked state is maintained.

[0052] In addition, the encrypted image identification server 120 can also determine by itself whether to register or identify a driver. For example, after the encrypted image identification server 120 receives the encrypted features, the encrypted image identification module 122 determines according to the comparison result with the registered encrypted features in the registered driver database 130. If the registered driver database 130 contains the corresponding registered encrypted features, the vehicle function is directly unlocked; otherwise, the vehicle function unlocking is refused, and a registration process is performed for this unregistered driver on the premise that it is confirmed that the unregistered driver has the registration permission.

[0053] In addition, the encrypted image recognition server 120 and the registered driver database 130 can be offline servers installed in the vehicle, or cloud servers connected through communication, or servers on the service provider side. Considering privacy issues, the offline server of the vehicle can be selected to execute this method. If it is considered that the performance of the offline server of the vehicle cannot handle the model operation requirements, it can be transmitted through network connection to the cloud server or the service provider side server for operation and wait for the result to be returned.

[0054] Figure 2 Schematic diagram of the driver registration process 200 in the driver face image encryption recognition method according to an embodiment of the present disclosure. The following refers to Figure 1 and Figure 2 to illustrate the process of a driver registering with the driver face image encryption recognition system 100.

[0055] When a driver first uses the driver face image encryption recognition system 100 and needs to register as a registered driver, and stores relevant data in the registered driver database 130, therefore, first step S201 is performed. The driver faces the display 110, so that the reflected light L1 reflected by the face F passes through the light-transmitting plate 111 and the light-transmitting display panel 112, and then is received by the optical encryption element 113 and forms encrypted light L2 through a plurality of light-transmitting holes H, and step S202 is performed. The encrypted light L2 is received by the image extraction device 114. The encrypted light L2 forms an encrypted image EI in the image extraction device 114, and then is transmitted to the encrypted image recognition server 120.

[0056] After receiving the encrypted image EI, the encrypted image recognition server 120 performs step S203 to extract the encrypted features of the encrypted image EI with the feature extraction module 121, and then performs step S204 to store the encrypted features in the registered driver database 130. At this time, the registered driver becomes a registered driver, and the corresponding encrypted features are registered encrypted features. Here, the feature extraction module 121 extracts the feature vector of the encrypted image EI as the encrypted features with the feature extraction model and stores them. And the encrypted features so far are only the features of the encrypted image EI, rather than the features of the face image FI, so that the actual image of the driver's face F can be avoided from leaking out.

[0057] Figure 3 Schematic diagram of the driver recognition process 300 in the driver face image encryption recognition method according to an embodiment of the present disclosure. The following refers to Figure 1 and Figure 3 to illustrate the process of a driver recognizing with the driver face image encryption recognition system 100.

[0058] When a driver wants to use the vehicle, the driver needs to first perform driver identification on the vehicle using the driver face image encryption identification system 100 to verify whether the driver is a registered driver. If so, the vehicle functions can be unlocked and the vehicle can be started. First, step S301 is performed. The driver faces the display 110, so that the reflected light L1 reflected by the face F passes through the light-transmitting plate 111 and the light-transmitting display panel 112, and then is received by the optical encryption element 113 and forms encrypted light L2 through a plurality of light-transmitting holes H. Then, step S302 is performed. The encrypted light L2 is received by the image extraction device 114. The encrypted light L2 forms an encrypted image EI in the image extraction device 114 and is transmitted to the encrypted image identification server 120.

[0059] After the encrypted image identification server 120 receives the encrypted image EI, step S303 is performed to extract the feature vector of the encrypted image EI as the encrypted feature by the feature extraction module 121. Then, step S304 is performed. The encrypted image identification module 122 accesses the registered driver database 130 and compares the encrypted feature with the registered encrypted features of the registered drivers stored in the registered driver database 130, and determines whether this encrypted feature is the registered encrypted feature of a registered driver according to the similarity of the feature vectors of the encrypted feature and the registered encrypted features (corresponding to step S305). After the judgment is completed, the encrypted image identification module 122 can further generate identification result data including the judgment result and transmit it to the vehicle locking module 123. The vehicle locking module 123 determines whether to unlock the vehicle functions for this driver according to the identification result data. If the identification result data shows that the encrypted feature is the registered encrypted feature of a registered driver, step S306 is performed to unlock the vehicle functions. On the contrary, if the identification result data shows that the encrypted feature is not the registered encrypted feature of a registered driver, step S307 is performed to reject unlocking the vehicle functions and maintain the locked state.

[0060] In an embodiment, whether the driver registration is successful, whether the driver identification result is successful, and whether the vehicle is unlocked can all be displayed to the driver by the encrypted image identification server 120 transmitting relevant information to the display 110.

[0061] Although the present disclosure has been disclosed as above with embodiments, it is not intended to limit the present disclosure. Those skilled in the art can make some changes and modifications without departing from the spirit and scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be subject to that defined by the appended claims.

Claims

1. A driving facial image encryption recognition system, comprising: A display, arranged in a direction facing a driver's face, the display comprising: An optical encryption element having a plurality of light-transmitting holes, the optical encryption element is used to receive a reflected light reflected from the driver's face to form an encrypted light through the plurality of light-transmitting holes; as well as an image capturing device for receiving the encrypted light and forming an encrypted image; and an encrypted image recognition server, connected to the image capture device for receiving the encrypted image, the encrypted image recognition server comprising: an encrypted image feature extraction module, used to extract an encrypted feature of the encrypted image; as well as An encrypted image recognition module is used to compare the encrypted feature with a registered encrypted feature to determine whether the encrypted feature is a registered encrypted feature of a registered driver, and generate a recognition result data accordingly.

2. The driving facial image encryption recognition system as claimed in claim 1, wherein the encrypted image recognition server further comprises: A vehicle locking module is used to unlock a vehicle function when the identification result data shows that the encryption feature is the registered encryption feature.

3. The driving facial image encryption recognition system as claimed in claim 1, further comprising: a registered driver database, connected to the encrypted image recognition server for storing the encrypted feature corresponding to the face extracted by the encrypted image feature extraction module when the driver is registered as the registered driver as the registered encrypted feature; The encrypted image recognition module accesses the registered driving database to compare the encrypted feature with the registered encrypted feature.

4. The driving facial image encryption recognition system as claimed in claim 1, wherein the display further comprises: A light-transmitting display panel is disposed between the optical encryption element and the driver's face so that the reflected light passes through and is then received by the optical encryption element. The light-transmitting display panel is connected to the encrypted image recognition server for communication and is used to receive the recognition result data and display a corresponding recognition result.

5. The driving facial image encryption recognition system as claimed in claim 1, wherein the display further comprises: An optical element is used to emit a light to the face of the driver, the light is visible light or infrared light, and the light reflects the reflected light after irradiating the face.

6. The driver's facial image encryption recognition system as claimed in claim 1, wherein a first distance between the driver's face and the optical encryption element is at least 150 times a second distance between the optical encryption element and the image capture device.

7. The driving facial image encrypted recognition system as described in claim 1, wherein the encrypted image recognition module compares the encrypted feature with the registered encrypted feature, calculates a similarity according to the distance between the feature vectors of the encrypted feature and the registered encrypted feature, and determines whether the encrypted feature is the registered encrypted feature according to whether the similarity is higher than a preset value.

8. A method for encrypted recognition of a driver's facial image, comprising: Receiving a reflected light reflected from a driver's face so that the reflected light passes through a plurality of light-transmitting holes of an optical encryption element to form an encrypted light; Making the encrypted light form an encrypted image; extracting an encryption feature of the encrypted image; as well as The encrypted feature is compared with a registered encrypted feature to determine whether the encrypted feature is a registered encrypted feature of a registered driver, and an identification result data is generated accordingly.

9. The method for encrypted recognition of a driver's facial image as claimed in claim 8, wherein the registered encrypted feature is the encrypted feature corresponding to the face extracted when the driver is registered as the registered driver.

10. The method for encrypted recognition of driving facial images as claimed in claim 8, further comprising: A vehicle function is unlocked when the identification result data shows that the encryption feature is the registered encryption feature.

Images