Face strong password

Abstract

The invention discloses a face strong password, which comprises the following steps of: starting an instruction, identifying a face specific action instruction, and carrying out next password identification only when the identification is successful; face data matching is carried out, face data are recognized, and next password recognition can be carried out after recognition succeeds; and the instruction is ended, the specific action instruction of the human face is recognized, if the recognition succeeds, the strong password of the human face is successfully matched, and the operability of related services is unlocked. According to the method, verification of the starting instruction is added before face recognition and verification of the ending instruction is added after face recognition, so that the strength of a face password is improved, meanwhile, verification instructions of the starting instruction and the ending instruction can be changed due to the fact that verification of the starting instruction and verification of the ending instruction are both achieved through face actions, and the verification efficiency is improved. And the protection performance of the face password is higher.

Description

Technical Field

[0001] The present invention relates to the field of face recognition technology, and in particular to a strong face password. Background Art

[0002] Face recognition is a biometric technology that identifies people based on their facial features. It uses a camera or a webcam to collect images or video streams containing faces, and automatically detects and tracks faces in the images, and then performs a series of face-related technologies on the detected faces. It is usually also called portrait recognition or facial recognition. Nowadays, face recognition is often used as a password to execute the switch passwords of electronic devices such as mobile phone screens, electronic payments, and electronic gates.

[0003] After the existing face recognition is entered, the generated password is unique and cannot be changed, unlike a digital password that can be replaced with a new password. Everyone's face is unique, and different face recognition algorithms can be misused in some way to a certain extent. For example, some algorithms can verify the password through a face picture without the need for real-life recognition. Or, the person who needs the face password may be pointed at by a hidden mobile phone camera without knowing it, and the face recognition is quietly completed. Once a person's portrait is stolen or secretly photographed, the password is temporarily leaked, and the consequences are self-evident. Therefore, the face recognition passwords currently used have the defects of low security protection and easy leakage. For this reason, a strong face password is proposed. Summary of the invention

[0004] The purpose of the present invention is to provide a strong face password to solve the problems raised in the above background technology.

[0005] To achieve the above purpose, the present invention provides the following technical solution: a strong face password, comprising: Open the command to identify the specific facial action commands of the face. Only after successful recognition can the next step of password recognition be carried out; Face data matching, face data recognition, the next step of password recognition can only be carried out after successful recognition; End the instruction and identify the specific facial action instructions. If the recognition is successful, the facial strong password is matched successfully, unlocking the operability of related services.

[0006] Preferably, the verification of the face strong password specifically includes the following steps: Step 1: Identify the opening command, verify the specific action command made by the face with the opening command data stored in the database, and proceed to the next step if the match is successful, otherwise it needs to be re-matched if the match fails; Step 2: After the start instruction in step 1 is successfully recognized, face recognition is performed, and the face data of the person is verified with the face data stored in the database. If the match is successful, the next step is entered. If the match fails, re-matching is required; Step three, after the face recognition in step two is successful, enter the end instruction recognition, and match the end instruction action made on the face with the end instruction data stored in the database. If the match is successful, the strong password is matched successfully and the relevant business is opened. If the match fails, it needs to be re-matched.

[0007] Preferably, the process of entering the identification password of the opening instruction specifically includes the following steps: Turn on the camera and enter the command input procedure; The face of the person makes facial movements that can unlock the door; The camera records the facial action command and stores the facial action command of the end opening command in a database.

[0008] Preferably, the process of entering the facial data of the face recognition comprises the following steps: Turn on the camera and enter the face recognition registration procedure; The face recognition program enters the facial data of the face through the camera and stores the facial data in the database.

[0009] Preferably, in step 3, the process of entering the end instruction specifically includes the following steps: Turn on the camera and enter the end command entry procedure; The face of the person makes a facial action command to end the unlocking; The camera records the facial motion command and stores the facial motion command of the unlock end command in a database.

[0010] Preferably, the verification passwords of the start instruction and the end instruction are face movement instructions, and the face movement instructions in the start instruction and the end instruction are modifiable.

[0011] Preferably, the facial action instructions for start and end instruction verification include but are not limited to blinking position and number, face turning direction and number, and / or mouth opening number. Preferably, during the verification process of the start instruction, face recognition and end instruction, if the instruction matching fails, the verification stage of the start instruction is returned to again.

[0012] Preferably, the face recognition algorithm used in the face strong password includes but is not limited to principal component analysis (PCA), linear discriminant analysis (LDA), Gaussian mixture model (GMM), support vector machine (SVM) or convolutional neural network (CNN).

[0013] Preferably, in step three, the related businesses executed include but are not limited to electronic payment, opening of safe access, screen unlocking and financial transfer.

[0014] Technical effects and advantages of the present invention: The present invention not only improves the strength of the face password by adding verification of the start command before face recognition and verification of the end command after face recognition, but also makes the verification of the start command and the end command changeable because both the verification of the start command and the end command are verified by facial movements, and the protection of the face password is higher; The face strong password composed of start command verification, face recognition verification and end command verification proposed by the present invention can solve the risk of easy impersonation and password leakage in the single verification of traditional face recognition, making face verification safer and more complete; The present invention adds corresponding facial movements before and after face recognition verification to match and verify the start instruction verification and the end instruction verification of unlocking, which can prevent the facial features of the user from being misused without the user's knowledge, thereby enhancing the protection of face recognition as a verification password. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The figure is a schematic diagram of the process of entering a strong password according to the present invention.

[0016] Figure 2 Schematic diagram of the verification process of the strong password of the present invention. DETAILED DESCRIPTION

[0017] 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 described embodiments 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.

[0018] The present invention provides Figure 1-2The face strong password shown includes: an opening instruction, which recognizes the face-specific action instructions, and the next step of password recognition can only be carried out after successful recognition. During the verification process of the opening instruction, it is necessary to record the specific actions made by the face through the camera, and match the recorded image with the image of the unlocking and opening instruction stored in the database. If the match is successful, the unlocking of the opening instruction is realized; face data matching, which recognizes the face data of the face, and the next step of password recognition can only be carried out after successful recognition. In face data matching, since the face data of each person is unique, it is only necessary to record the face data image into the system through the camera and compare it with the face image stored in the database. If the match is successful, the face data matching is unlocked; an end instruction, which recognizes the face-specific action instructions, and if the recognition is successful, the face strong password is matched successfully, and the operability of the unlocking related business is realized. The end instruction and the opening instruction are originally The principle is similar. It is necessary to record the facial movements through the camera, and then match the recorded image with the verification image of the end instruction stored in the system's database. If the match is successful, the end instruction verification is successful, and the entire face strong password is enabled. The verification passwords of the enable instruction and the end instruction are face movement instructions, and the face movement instructions in the enable instruction and the end instruction are changeable. Since the verification of the enable instruction and the end instruction uses face movement instructions for verification, compared with the uniqueness of facial recognition, the facial movement verification in the enable instruction and the end instruction can be changed. Therefore, in the face strong password, the verification password of the enable instruction and the verification password of the end instruction can be replaced, so that it can be changed through the face strong password, making the face password more secure and not easy to be impersonated. At the same time, it is not easy for others to successfully verify with photos, making facial recognition more secure. Embodiment 1

[0019] The verification of the face strong password specifically includes the following steps: Step 1: Turn on the camera of the device, turn on the recognition of the command, point the face of the person at the camera, and let the camera record the specific action command made by the face of the person, and then verify the recorded image data with the opening command data stored in the database. If the match is successful, proceed to the next step. If the match fails, re-match is required; Step 2: After the start command in step 1 is successfully recognized, the face is aligned with the camera again as in step 1, but no action is required on the face, so that the camera records the face image and performs face recognition. The face data is verified with the face data stored in the database. If the match is successful, the next step is entered. If the match fails, the match needs to be repeated. Step three, after the face recognition in step two is successful, continue to align the face with the camera and make facial movements of unlocking end instructions on the face, so that the camera will record the facial movements and enter the recognition of the end instruction. The end instruction movement made on the face is matched with the end instruction data stored in the database. If the match is successful, the strong password is matched successfully and the related business is opened. If the match fails, it needs to be re-matched. The related businesses to be executed include but are not limited to electronic payment, opening of safe access control, screen unlocking and financial transfer.

[0020] The process of adding start command verification and end command verification before and after face recognition is equivalent to adding front and back valves to the face recognition password verification, making the usage rights more protective. The verification of the start command and the end command is to insert the face capture features before and after the face recognition verification, combined with the changeable facial movement features, the traditional face recognition password can be changed, thereby enhancing the protection and security of the face password.

[0021] It should be noted that: in the above example, during the verification process of the start command, face recognition and end command, if the command matching fails, it will return to the verification stage of the start command. In this way, the verifier needs to return to the initial position for re-verification after making mistakes one by one, which makes the verification more cumbersome. It can effectively avoid the disadvantages of successful verification by taking shortcuts, and make the use of strong facial passwords more protective. Embodiment 2

[0022] The first step is to enter the identification password of the opening instruction, which specifically includes the following steps: Turn on the camera and enter the command input procedure; The face of the person makes a facial action command that can unlock the device. The facial action command can be the order and number of blinking of the left and right eyes, the number of times of opening the mouth, and the direction and number of shaking the head, or a combination of various previous commands; The camera records the facial action command and stores the facial action command of the end opening command in a database.

[0023] The second step is to input the facial data of the face recognition, which specifically includes the following steps: Turn on the camera and enter the face recognition registration procedure; The face recognition program enters the facial data of the face through the camera and stores the facial data in the database.

[0024] The process of entering the end instruction specifically includes the following steps: Step 3: Turn on the camera and enter the end command entry procedure; The face of the person performs a facial action command to end the unlocking, which may be the order and number of blinking of the left and right eyes, the number of times of opening the mouth, and the direction and number of times of shaking the head, or a combination of various previous commands; The camera records the facial motion command and stores the facial motion command of the unlock end command in a database.

[0025] Furthermore, facial motion instructions for verifying the start and end instructions include but are not limited to the position and number of blinks, the direction and number of facial turns, and / or the number of mouth openings. Since the position and number of blinks are uncertain, the direction and number of facial turns are also random, and the number of mouth openings is also unpredictable, using these facial motions as verification passwords for start and end instructions can make the verification passwords for start and end instructions exponentially more difficult, and the verification passwords for start and end instructions can be changed by changing the types, sequences, and numbers of these motions. In this way, after the start and end instructions are accidentally leaked, they can be replaced. Compared with the uniqueness of face recognition, the use of strong face passwords can be more secure and concealed. The face recognition algorithms used in strong face passwords include but are not limited to principal component analysis (PCA), linear discriminant analysis (LDA), Gaussian mixture model (GMM), support vector machine (SVM) or convolutional neural network (CNN). Among them, principal component analysis (PCA) is based on statistical principles to convert face images into a set of principal component features, and then identify them by comparing feature vectors; linear discriminant analysis (LDA) is similar to PCA. LDA is also an algorithm based on statistical methods. It extracts and classifies features by maximizing the divergence between classes and minimizing the divergence within classes; Gaussian mixture model (GMM ... identifies them by comparing feature vectors; linear discriminant analysis (LDA) is based on statistical methods to extract and classify features by maximizing the divergence between classes and minimizing the divergence within classes; Gaussian mixture model (GMM) is based on statistical principles to convert face images into a set of principal component features, and then identifies them by comparing feature vectors; linear discriminant analysis (LDA) is based on statistical principles to extract and classify features by maximizing the divergence between classes and minimizing the divergence within classes; Gaussian mixture model (GMM) is based on statistical principles to convert face images into a set of principal component features, and then identifies them by comparing feature vectors; linear discriminant analysis (LDA) is based on statistical principles to extract and classify features by maximizing the divergence between classes and minimizing the divergence within classes; Gaussian mixture model (GMM) is based on statistical principles to convert face images into a The model (GMM) uses Gaussian distribution to model the facial features of each person, and then recognizes them by comparing the similarities between the models; the support vector machine (SVM) projects the face image into a high-dimensional space and classifies it by constructing the optimal separation hyperplane in the space; the convolutional neural network (CNN) automatically learns the abstract features of the face through multi-layer convolution and pooling operations, and classifies it through the fully connected layer; in addition to the common face recognition algorithms mentioned above, you can also apply, for example, Eigenfaces is a face detection and recognition method used to determine the variance of faces in image data sets. It uses these differences to encode and decode faces through machine learning. A set of eigenfaces is a collection of "standardized face components" determined by statistical analysis of a large number of face images. Facial features are assigned mathematical values. Because this method does not use digital pictures, but statistical databases, any face is a combination of these values ​​in different percentages; skin recognition technology has many applications - face detection algorithms, bad image filtering, gesture analysis, etc. It usually uses high-resolution images, and special cases of skin texture analysis use different unique parameters, such as moles, skin color, skin color, etc.

[0026] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. 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 strong face password, characterized in that: include: Open the command to identify the specific facial action commands of the face. Only after successful recognition can the next step of password recognition be carried out; Face data matching, face data recognition, the next step of password recognition can only be carried out after successful recognition; End the instruction and identify the specific facial action instructions. If the recognition is successful, the facial strong password is matched successfully, unlocking the operability of related services.

2. A strong face password according to claim 1, characterized in that: The verification of the face strong password specifically includes the following steps Step 1: Identify the opening command, verify the specific action command made by the face with the opening command data stored in the database, and proceed to the next step if the match is successful, otherwise it needs to be re-matched if the match fails; Step 2: After the start instruction in step 1 is successfully recognized, face recognition is performed, and the face data of the person is verified with the face data stored in the database. If the match is successful, the next step is entered. If the match fails, re-matching is required; Step three, after the face recognition in step two is successful, enter the end instruction recognition, and match the end instruction action made on the face with the end instruction data stored in the database. If the match is successful, the strong password is matched successfully and the relevant business is opened. If the match fails, it needs to be re-matched.

3. A strong face password according to claim 2, characterized in that: In the step 1, the process of entering the identification password of the opening instruction specifically includes the following steps: Turn on the camera and enter the command input procedure; The face of the person makes facial movements that can unlock the door; The camera records the facial action command and stores the facial action command of the end opening command in a database.

4. A strong face password according to claim 2, characterized in that: In the step 2, the process of entering the facial data of the face recognition comprises the following steps: Turn on the camera and enter the face recognition registration procedure; The face recognition program records the facial data of the face through the camera and stores the facial data in the database.

5. A strong face password according to claim 2, characterized in that: In step 3, the process of entering the end instruction specifically includes the following steps: Turn on the camera and enter the end command entry procedure; The face of the person makes a facial action command to end the unlocking; The camera records the facial motion command and stores the facial motion command of the unlock end command in a database.

6. A strong face password according to claim 1, characterized in that: The verification passwords of the start instruction and the end instruction are face motion instructions, and the face motion instructions in the start instruction and the end instruction are modifiable.

7. A strong face password according to claim 6, characterized in that: The facial action instructions for start instruction and end instruction verification include but are not limited to blinking position and number of times, the direction and number of times the face turns and / or the number of times the mouth opens.

8. A strong face password according to claim 2, characterized in that: During the verification process of the start command, face recognition and end command, if the command matching fails, the verification stage of the start command will be returned to again.

9. A strong face password according to claim 2, characterized in that: The face recognition algorithm used in the face strong password includes but is not limited to principal component analysis (PCA), linear discriminant analysis (LDA), Gaussian mixture model (GMM), support vector machine (SVM) or convolutional neural network (CNN).

10. A strong face password according to claim 2, characterized in that: In step three, the related businesses performed include but are not limited to electronic payment, opening of the safe access control, screen unlocking and financial transfer.

Images