Intelligent face recognition access control method and system and storage medium thereof
Automatically identify and enter visitor facial data through the intelligent access control system, the problem of cumbersome facial recognition operation is solved, the degree of intelligence and user experience is improved, security and accuracy are enhanced, and data storage costs and misjudgment risks are reduced. .
Patent Information
- Application Number
- CN202411438500.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-15
- Publication Date
- 2025-01-10
AI Technical Summary
The existing intelligent access control system is complicated to handle the recognition of visitors' facial images, especially when the owner's relatives visit, the doorman needs to make multiple confirmations and inputs, which leads to cumbersome and time-consuming process.
By obtaining the guest facial recognition data and comparing it with the owner facial database, we will automatically judge the identity of relatives, generate a notification signal and output it. After the owner confirms, the guest facial data will be automatically entered. There is no need for a doorman to operate, and the facial muscles and expression recognition is combined to improve accuracy Sex and safety.
It realizes automatic identification and entry of visitors, improves intelligence and user experience, reduces data storage needs, reduces costs, improves security and accuracy, and reduces the probability of misjudgment and bad intentions of visitors entering.
Smart Images
Figure CN119296217A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of intelligent access control, and in particular to a facial recognition intelligent access control method, system and storage medium thereof. Background Art
[0002] With the increasing popularity of intelligent civil applications, more and more people are using smart door locks and smart homes, and more and more smart access control systems with facial recognition are being added at the entrances of residential communities, thus avoiding the situation where owners return home only to find that they have forgotten to bring their door cards and are unable to enter the community.
[0003] However, if a visitor comes to visit when the owner is at home, the security guard and the owner must first confirm the situation, and then the security guard is usually required to perform facial recognition and entry of the visitor, which is a cumbersome operation. Summary of the invention
[0004] In order to improve the problem that visitors need the doorman to operate facial recognition and input, which is cumbersome, the present application provides a facial recognition intelligent access control method, system and storage medium thereof.
[0005] This application provides a facial recognition intelligent access control method, which adopts the following technical solutions: A facial recognition intelligent access control method, comprising: Obtain visitor facial recognition data; Determining facial similarity through the visitor facial recognition data and a preset owner facial database; Determine the facial data of the relative owner through the facial similarity and a preset relative similarity threshold; Determine and output a notification signal through the relative owner's facial data and the owner's facial database; Get the owner's pass signal; The visitor's facial data is determined by the owner's passing signal, the relative owner's facial data and the facial similarity, and stored in a preset visitor's facial database.
[0006] By adopting the above technical solution, when the owner's relatives visit, there is no need for the doorman to operate. The relatives can directly identify and then automatically identify which owner's relative the visitor is and automatically notify the owner. Once the owner confirms the visit, the visitor's facial data is automatically entered. There is no need for the doorman to notify the owner or operate, which improves the level of intelligence and user experience.
[0007] Optional, including: The facial similarity includes facial features similarity and overall similarity, and the visitor facial recognition data includes visitor facial features recognition data and visitor overall recognition data; Determine the facial features similarity and facial features preference data through the facial features recognition data of the visitor and the facial features data of the relative or owner; Determine the overall similarity and overall preference data through the visitor's facial features recognition data, the visitor's overall recognition data and the relative owner's facial image data; Determine owner-related data through the relative owner's facial data and the owner's facial database; The visitor facial data is determined by the facial features similarity, the facial features preference data, the overall similarity, the overall preference data and the owner association data, and stored in the visitor facial data database.
[0008] By adopting the above technical solution, the memory occupied by each visitor's data in the visitor profile database is greatly reduced, the cost is reduced, and more visitor data can be stored, reducing the probability of some visitor data being cleared due to insufficient memory as time passes, thereby improving the user experience.
[0009] Optional, including: Acquiring time data, wherein the facial image similarity includes facial muscle similarity; Determine facial change data using the time data and the visitor facial recognition data; Determining the visitor's facial muscle data through the facial change data, a preset facial recognition threshold and the visitor's facial recognition data; The facial muscle similarity is determined by comparing the visitor's facial muscle data with the owner's facial image database.
[0010] By adopting the above technical solution, the facial muscles of visitors can be automatically identified through changes, and the difference between the visitor's face and the owner's face can be judged more accurately through facial muscle data, thereby improving accuracy and reducing the probability of misjudgment of non-relatives.
[0011] Optional, including: Determining visitor expression data through the visitor's facial muscle data, the visitor's facial feature recognition data and a preset expression threshold; A reminder signal is determined and outputted based on the visitor's facial expression data and a preset friendly expression threshold.
[0012] By adopting the above technical solution, whether the visitor has good intentions can be identified through facial expressions, and owners who are inconvenient to view images can be reminded, thereby reducing the probability of malicious relatives entering the community and protecting the owners.
[0013] Optional, including: Obtain visitor monitoring data; Forming a visitor monitoring database through the visitor monitoring data and the time data; Retrieving the required visitor monitoring data from the visitor monitoring database through the visitor facial recognition data; Determining hazard data by comparing the visitor monitoring data with a preset hazard threshold; A reminder signal is determined and outputted based on the hazard data, the visitor expression data, the benevolent expression and a preset hazard deviation threshold.
[0014] By adopting the above technical solution, it is automatically determined whether the visitor is carrying highly dangerous items. At the same time, combined with the visitor's expression, a comprehensive judgment is made as to whether the visitor is carrying malicious intentions, so as to protect the safety of the owner and reduce the probability of the owner receiving deliberate retaliation.
[0015] The present application provides a facial recognition intelligent access control system, which adopts the following technical solutions: A facial recognition intelligent access control system, comprising: A facial recognition module is used to perform facial recognition data on the user, obtain the visitor facial recognition data and output it; The data processing module receives the visitor face recognition data, performs data processing and calculation, and outputs a corresponding notification signal or reminder signal.
[0016] By adopting the above technical solution, identification and judgment can be automatically realized, which is convenient and fast and improves the degree of intelligence.
[0017] Optionally, a plurality of wide-area remote monitoring cameras are included, and the wide-area remote monitoring cameras are used for monitoring, obtaining visitor monitoring data, and outputting the data to the data processing module.
[0018] By adopting the above technical solution, through the cross-coverage of multiple wide-area remote monitoring cameras, more distant and comprehensive monitoring can be achieved, reducing the probability of visitors hiding dangerous items.
[0019] The present application provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium stores a computer program that can be loaded by a processor and execute a facial recognition intelligent access control method.
[0020] By adopting the above technical solution, the computer program is stored in a computer-readable storage medium.
[0021] In summary, the present application includes at least one of the following beneficial technical effects: 1. Automatically input the visitor's facial data, without the need for the security guard to notify the owner or operate, which improves the level of intelligence and user experience.
[0022] 2. Greatly reduce the memory occupied by each visitor's data in the visitor profile database, reduce costs, and enable the storage of more visitor data, reduce the probability of some visitor data being cleared due to insufficient memory over time, and improve user experience. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 It is a flow chart of a facial recognition intelligent access control method in an embodiment of the present application.
[0024] Figure 2 It is a flowchart diagram of step S2 to step S24.
[0025] Figure 3 It is a flowchart diagram of step S3 to step S41.
[0026] Figure 4 It is a flowchart diagram of step S5 to step S54.
[0027] Figure 5 It is a module schematic diagram of a facial recognition intelligent access control system in an embodiment of the present application.
[0028] Explanation of the accompanying drawings: 1. Face recognition module; 2. Data processing module; 3. Wide-area remote monitoring camera. DETAILED DESCRIPTION
[0029] The following is combined with Figure 1-5 This application is described in further detail.
[0030] The present application embodiment discloses a facial recognition intelligent access control method. Figure 1 , the face recognition intelligent access control method includes the following steps: S1. Obtain visitor facial recognition data; S11, determining facial similarity through visitor facial recognition data and a preset owner facial database; S12, determining the facial data of the relative owner through the facial similarity and the preset relative similarity threshold; S13, determining a notification signal through the relative owner's facial image data and the owner's facial image database and outputting it; S14, obtaining a signal from the owner passing through; S15, determining the visitor's facial data through the owner's passing signal, the facial data of the relative owner and the facial similarity, and storing it in a preset visitor's facial database.
[0031] In detail: the visitor facial recognition data is obtained by taking pictures with devices such as high-precision cameras, and can be images or videos, with videos being the preferred choice; the owner facial recognition database is a database that records the facial recognition data and owner information (including the owner's personal information, information about the real estate owned by the owner, etc.) of all owners. The visitor facial recognition data is compared with the data in the owner facial recognition database to obtain similarities in turn, and the highest similarity is taken as the facial similarity. If the facial similarity exceeds the relative similarity threshold, it can be determined that the current visitor is a relative of the corresponding owner, and the facial recognition data of this owner is determined to be the facial data of the relative owner. At this time, the owner's information is obtained through the owner information in the owner facial recognition database, and the corresponding notification signal is output to the owner, such as notifying the access control system in the owner's house, or sending information to the owner's terminal app, or sending a text message to the owner's mobile phone, etc. After the owner chooses to pass, the owner will send an owner pass signal to the access control system at the entrance of the community. At this time, the visitor facial data will be generated and stored in the visitor facial database for the visitor's next visit.
[0032] Reference Figure 2 , when generating visitor profile data, the following steps are included: S2. Face similarity includes facial features similarity and overall similarity. Visitor face recognition data includes visitor facial features recognition data and visitor overall recognition data. S21, determining facial features similarity and facial features preference data through the facial features recognition data of the visitor and the facial features data of the relative or owner; S22, determining overall similarity and overall preference data through visitor facial recognition data, visitor overall recognition data and relative owner facial image data; S23, determining owner-related data through the facial image data of the relative owner and the facial image database of the owner; S24, determining the visitor's facial image data through facial feature similarity, facial feature preference data, overall similarity, overall preference data and owner-related data, and storing it in the visitor's facial image database.
[0033] Details: Visitor facial feature recognition data is the recognition data obtained after recognizing the visitor's facial features. Visitor overall recognition data is the data after removing the facial features, such as the distance between the eyes and the nose, the size of the forehead, the width of the cheeks, the distance between the nose and the mouth, the length of the chin, etc. Visitor facial feature recognition data combined with visitor overall recognition data is the data of the visitor's entire face. The visitor facial feature recognition data is compared with the relative owner's facial data to obtain the similarity of the facial features, which is the facial feature similarity. If the eyes in the visitor's facial feature recognition data are more upturned than the eyes in the relative owner's facial data, this "More hanging" is the facial feature preference data to indicate the difference between the visitor and the relative owner. The specific data can be obtained through the ratio of the hanging amplitude to the owner. For example, the hanging angle of the eye end of the relative owner's facial data is 10°, while the hanging angle of the eye end of the visitor's facial feature recognition data is 12°. The facial feature preference data can be recorded as (hanging eye end, 20%). The overall similarity is the comparison between the visitor's overall recognition data and the relative owner's facial data. If the visitor's overall preference data needs to be determined, it is also necessary to refer to the facial features in the visitor's facial feature recognition data, so as to obtain, for example, the visitor's forehead to eyebrows. The distance from the center of the owner's forehead to the center of the eyebrows is 5cm, and the distance from the owner's forehead to the center of the eyebrows is 4.5cm. The overall preference data can be recorded as (forehead-brow distance, (5-4.5) / 4.5=11%), and the facial features similarity and overall similarity must both exceed the kinship similarity threshold to prevent the probability of misjudgment due to accidental similarity of partial facial features of non-relatives; the owner-related data is the number of the relative owner's facial data in the owner's facial database. For example, the owner x of Room 1, Building 1 is numbered 001, which is the owner-related data. When recorded in the visitor facial database, only 001 is recorded without storing the entire relative owner facial data. The facial similarity, facial preference data, overall similarity, overall preference data and owner-related data of the visitor are recorded, so as to achieve the purpose of not storing the facial recognition data of the visitor and the facial data of the relative owners, so as to reduce the memory usage. When the facial recognition data of the visitor needs to be viewed in the future, the image data of the owner can be read from the owner facial database through the owner-related data, and then the facial recognition data of the visitor can be reversely restored through the facial similarity and irrelevant preference data. The overall recognition data of the visitor is restored through the overall similarity, overall preference data and the restored facial recognition data of the visitor.
[0034] Reference Figure 3 , further comprising the following steps: S3, obtaining time data, face similarity including facial muscle similarity; S31, determining facial change data through time data and visitor facial recognition data; S32, determining the visitor's facial muscle data through the facial change data, the preset facial recognition threshold and the visitor's facial recognition data; S33, determining facial muscle similarity through the visitor's facial muscle data and the owner's facial image database.
[0035] In detail: Through the visitor facial recognition data in different time data, such as the eyelids of the visitor when he lowers his eyebrows and looks downward, and the eyelids of the visitor when he raises his eyes and looks up, through the eyelids in these two different states, combined with the facial recognition threshold, the shape and distribution of the eyelid muscles can be determined, and the two eyelids in different states are the facial change data, and the obtained eyelid muscles are the visitor's facial muscle data, which are marked in the visitor facial recognition data, and then compared with the data of each owner in the owner facial database to determine the similarity, and the highest one is taken as the facial muscle similarity, and then compared with the relative similarity threshold.
[0036] Reference Figure 3 , further comprising the following steps: S4, determining the visitor's expression data through the visitor's facial muscle data, the visitor's facial features recognition data and a preset expression threshold; S41, determining a reminder signal based on the visitor's facial expression data and a preset friendly expression threshold and outputting the reminder signal.
[0037] In detail: suppose the expression threshold is (a, b, c, d, e, f). For example, smiling uses multiple facial muscles such as the orbicularis oculi, zygomaticus major, zygomaticus minor, levator labii superioris, levator anguli oris, and risorius muscle, and the expression threshold is the coefficient of the expression that can be produced by the amplitude of each muscle movement. Suppose the orbicularis oculi, zygomaticus major, zygomaticus minor, levator labii superioris, levator anguli oris, and risorius muscle are stretched or contracted by the amplitudes of g, h, i, j, k, and l respectively, then the visitor's expression data can be obtained as a*g+b*h+c*i+d*j+e*k+f*l, and then compared with the benevolent expression threshold. If it is judged to be within the range of the benevolent expression threshold, then the visitor is judged to be benevolent at this time. If it is outside the range of the benevolent expression threshold, then the visitor is judged to be not benevolent at this time, but not necessarily malicious, for example, expressionless, then the corresponding reminder signal is output to the owner at this time. The owner may not be able to view the visitor's expression through the access control system at home, so a reminder is needed.
[0038] Reference Figure 4 , further comprising the following steps: S5. Obtain visitor monitoring data; S51, forming a visitor monitoring database through visitor monitoring data and time data; S52, calling the required visitor monitoring data from the visitor monitoring database through the visitor face recognition data; S53, determining hazard data through visitor monitoring data and preset hazard thresholds; S54, determining and outputting a reminder signal based on the hazard data, visitor expression data, benevolent expressions and a preset hazard deviation threshold.
[0039] Details: The visitor monitoring data is captured by a camera with a larger range. The shooting accuracy may be lower than the shooting accuracy required for visitor facial recognition. The visitor monitoring data is stored in the visitor monitoring database in combination with the time data. When the visitor facial recognition data recognizes the visitor's face and starts working, the corresponding visitor monitoring data is called from the visitor monitoring database. For example, if the time data is 1 minute, and the visitor enters the range of the visitor monitoring data for the first time, when the time data is 1 minute and 30 seconds, the corresponding visitor facial recognition data is recognized and the data processing work is started from the owner facial database. At this time, the visitor monitoring data between 1 minute and 30 seconds and 1 minute will be called from the visitor monitoring database. Since the visitor monitoring data is captured by multiple cameras at different angles, To more comprehensively identify the items carried by visitors, for example, a knife placed in a packaging box in a transparent plastic bag, the degree of danger is the hazard threshold. Suppose this hazard threshold is 1, and the hazard threshold for a kitchen knife held directly by the visitor is 10. After reading the data of all identified items and the hazard threshold, add them up to get the hazard data, and then combine the hazard data with the hazard deviation threshold. For example, if the hazard data is 5, and the hazard deviation threshold is 0.1, if the visitor's expression data is 5.6 and the benevolent expression is 6, it means that the visitor's expression data is less than 6 for benevolence, and the smaller the data, the more benevolent it is. 5.6 means that the visitor has benevolence at this time, but the data corrected by combining the hazard data is 5.6+5*0.1=6.1, which exceeds the range of benevolence, and then an output indicating non-benevolence reminder signal is given to the owner.
[0040] The present application embodiment discloses a facial recognition intelligent access control system. Figure 5 The facial recognition intelligent access control system includes a facial recognition module 1, a data processing module 2, and a wide-area remote monitoring camera 3. The facial recognition module 1 can use a camera with high recognition accuracy. The number of the cameras can be only one and installed at the gate of the community to collect the user in front of the door to obtain the visitor facial recognition data. The facial data of the owner in the owner facial database can also be collected and entered by the facial recognition module 1. The wide-area remote monitoring camera 3 can use a camera with a lower recognition accuracy than the facial recognition module 1. The number of the wide-area remote monitoring cameras 3 is at least two and is distributed at different positions of the gate of the community, such as on both sides, to monitor the gate and the distance to obtain multiple visitor monitoring data.
[0041] Reference Figure 5The data processing module 2 includes a processor and a database. The database is used to store various threshold data, such as the owner's face database, relative similarity threshold, expression threshold, benevolent expression threshold, hazard threshold, hazard deviation threshold, visitor monitoring database and other threshold data. After the processor receives the visitor face recognition data and the visitor monitoring data, it calls the corresponding threshold data from the database for calculation to obtain the corresponding notification signal or reminder signal.
[0042] The processor may include a central processing unit such as a CPU or MPU or a host system built around a CPU or MPU, including hardware or software. After the meter has a processor, people can freely control the metering instrument through programming to make it run according to people's wishes. The processor can control local measurement transmission, remote measurement transmission, remote communication, etc. through internal protocols. Internal protocols refer to all protocols that achieve mutual communication or links within the same metering instrument or the same system, including: human-computer interaction protocols, software / hardware (interface) protocols, chip bus (C-Bus) protocols, internal bus (I-Bus) protocols, etc. Part or all of the protocols. With the development of integrated circuit technology, some protocols that belong to the external bus (E-Bus) protocol are also classified as internal protocols after the external bus (E-Bus) is integrated into the chip.
[0043] The embodiment of the present application discloses a computer-readable storage medium, which stores a computer program that can be loaded by a processor and execute a facial recognition intelligent access control method.
[0044] Computer-readable storage media include, for example, various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.
[0045] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.
Claims
1. A facial recognition intelligent access control method, characterized in that: include: Obtain visitor facial recognition data; Determining facial similarity through the visitor facial recognition data and a preset owner facial database; Determine the facial data of the relative owner through the facial similarity and a preset relative similarity threshold; Determine and output a notification signal through the relative owner's facial data and the owner's facial database; Get the owner's pass signal; The visitor's facial data is determined by the owner's passing signal, the relative owner's facial data and the facial similarity, and stored in a preset visitor's facial database.
2. According to claim 1, a facial recognition intelligent access control method is characterized in that: include: The facial similarity includes facial features similarity and overall similarity, and the visitor facial recognition data includes visitor facial features recognition data and visitor overall recognition data; Determine the facial features similarity and facial features preference data through the facial features recognition data of the visitor and the facial features data of the relative or owner; Determine the overall similarity and overall preference data through the visitor's facial features recognition data, the visitor's overall recognition data and the relative owner's facial image data; Determine owner-related data through the relative owner's facial data and the owner's facial database; The visitor facial data is determined by the facial features similarity, the facial features preference data, the overall similarity, the overall preference data and the owner association data, and stored in the visitor facial data database.
3. The facial recognition intelligent access control method according to claim 2, characterized in that: include: Acquiring time data, wherein the facial image similarity includes facial muscle similarity; Determine facial change data using the time data and the visitor facial recognition data; Determining the visitor's facial muscle data through the facial change data, a preset facial recognition threshold and the visitor's facial recognition data; The facial muscle similarity is determined by comparing the visitor's facial muscle data with the owner's facial image database.
4. The facial recognition intelligent access control method according to claim 3, characterized in that: include: Determining visitor expression data through the visitor's facial muscle data, the visitor's facial feature recognition data and a preset expression threshold; A reminder signal is determined and outputted based on the visitor's facial expression data and a preset friendly expression threshold.
5. The facial recognition intelligent access control method according to claim 4, characterized in that: include: Obtain visitor monitoring data; Forming a visitor monitoring database through the visitor monitoring data and the time data; Retrieving the required visitor monitoring data from the visitor monitoring database through the visitor facial recognition data; Determining hazard data by comparing the visitor monitoring data with a preset hazard threshold; A reminder signal is determined and outputted based on the hazard data, the visitor expression data, the benevolent expression and a preset hazard deviation threshold.
6. A facial recognition intelligent access control system, using the facial recognition intelligent access control method according to claim 5, characterized in that: include: A facial recognition module (1) is used to perform facial recognition data on a user, obtain the visitor facial recognition data and output it; The data processing module (2) receives the visitor face recognition data, performs data processing and calculation, and outputs a corresponding notification signal or reminder signal.
7. The facial recognition intelligent access control system according to claim 6, characterized in that: It comprises a plurality of wide-area remote monitoring cameras (3), wherein the wide-area remote monitoring cameras (3) are used for monitoring, obtaining visitor monitoring data, and outputting the data to the data processing module (2).
8. A computer-readable storage medium, characterized in that: A computer program is stored which can be loaded by a processor and executes a facial recognition intelligent access control method as claimed in any one of claims 1 to 5.