Intelligent access control face recognition equipment for park
By introducing face recognition simulation modules and algorithm optimization modules into the park access control equipment, the problem of unsatisfactory recognition effect in complex environments is solved, and the recognition accuracy and efficiency are improved. By regularly changing the recognition algorithms, we can adapt to environmental changes.
Patent Information
- Application Number
- CN202411410483.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-10
- Publication Date
- 2024-12-20
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing campus access control equipment has poor facial recognition effect in complex environments and cannot automatically adjust the recognition strategy, resulting in low recognition accuracy and efficiency.
A smart access control face recognition device in the park was designed, including a face recognition simulation module, a recognition algorithm simulation module and a recognition algorithm optimization module. By periodically generating a face recognition simulation environment, a variety of face recognition algorithms are set for simulation recognition. , analyze the recognition effect of each algorithm, select the preferred algorithm for face recognition, and regularly replace the recognition algorithm to ensure recognition accuracy and efficiency.
Through the simulation and analysis of a variety of face recognition algorithms, the algorithm with the best recognition capability is selected for recognition, which improves the recognition accuracy and efficiency of the park access control equipment in various environments, and ensures the stability and adaptability of the recognition effect.
Smart Images
Figure CN119169731A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of access control recognition technology, and more specifically, to a campus smart access control face recognition device. Background Art
[0002] With the rapid development of science and technology, especially driven by artificial intelligence and the Internet of Things, intelligent park management has become an important means to improve operational efficiency and enhance security. As the first line of defense for park security management, the level of intelligence of the access control system directly affects the overall security and management efficiency of the park.
[0003] In recent years, face recognition technology has gradually become the preferred technology for access control system upgrades due to its advantages such as non-contact, high accuracy, and difficulty in forging. At present, although face recognition technology is widely used in campus access control equipment, it is limited to one face recognition algorithm and cannot automatically adjust the recognition strategy according to environmental changes, resulting in unsatisfactory recognition results in some complex environments. Therefore, how to ensure the recognition effect of campus access control equipment in various environments is an urgent problem to be solved. Summary of the invention
[0004] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a campus smart access control face recognition device.
[0005] To achieve the above object, the present invention provides the following technical solutions:
[0006] A park smart access control face recognition device, including a face recognition simulation module, a recognition algorithm simulation module, and a recognition algorithm optimization module;
[0007] The face recognition simulation module is used to periodically generate i face recognition simulation environments;
[0008] The recognition algorithm simulation module sets each type of face recognition algorithm in each face recognition simulation environment to perform simulation recognition on the basic face photo in turn, thereby generating face algorithm recognition records of each type of face recognition algorithm;
[0009] After the recognition algorithm optimization module completes the simulation recognition in all face recognition simulation environments, the access control performs face recognition based on the optimization recognition algorithm.
[0010] Furthermore, i face recognition simulation environments are generated periodically, specifically: a time interval g is preset, and when the time node of the time interval g is reached, the environmental data around the access control is collected, i basic face photos are set synchronously, and i digital twin models are constructed based on the environmental data and the basic face photos, and the face recognition simulation environment is generated based on the digital twin model.
[0011] Furthermore, the face algorithm recognition record includes the face recognition simulation environment number, the face recognition algorithm type, and the algorithm analysis value.
[0012] Furthermore, the algorithm analysis value of the face recognition algorithm is obtained in the following manner: during the process of the face recognition algorithm identifying the basic face photo, each time a face feature point is identified, a face recognition analysis record of the face feature point is generated, a rectangular coordinate system is constructed with the face recognition analysis value as the Y-axis and the face recognition time as the X-axis, the point of the corresponding face recognition analysis value is marked in the rectangular coordinate system according to the recognition time, the points of two adjacent recognition times in the rectangular coordinate system are connected to obtain a face recognition analysis curve, the two endpoints of the face recognition analysis curve are perpendicular to the X-axis to obtain two start and end analysis line segments, and the total area of the figure formed by the face recognition analysis curve, the two start and end analysis line segments and the X-axis is marked as the algorithm analysis value.
[0013] Furthermore, the face recognition analysis record includes the face feature name, the face recognition analysis value, and the recognition time.
[0014] Furthermore, the face recognition analysis value of the face recognition analysis record is obtained in the following manner: when a facial feature point is identified, the feature data identified by the facial feature point is marked as real-recognition feature data, the feature data of the facial feature point in the basic face photo is extracted and marked as basic feature data, the real-recognition feature data and the basic feature data are pre-processed, the processed real-time feature data and the basic feature data are packaged into a feature data comparison set, the feature data comparison set is used as input data of the face recognition analysis model, and the face recognition analysis model outputs the face recognition analysis value.
[0015] Furthermore, all face algorithm recognition records are collected to obtain the algorithm recommendation value of each face recognition algorithm type, and the face recognition algorithm type with the largest algorithm recommendation value is marked as the preferred recognition algorithm.
[0016] Furthermore, the algorithm recommendation value of each face recognition algorithm type is obtained by the following method: obtaining the face recognition algorithm type of the face algorithm recognition record, marking the face algorithm recognition records of the same face recognition algorithm type as the same type of algorithm recognition records, summing up the algorithm analysis values of all the same type of algorithm recognition records and taking the average to obtain the algorithm analysis mean Lksz, grouping all the same type of algorithm recognition records into one group, summing up the algorithm analysis values of two same type of algorithm recognition records in the same group to obtain the algorithm analysis overlap value, setting the algorithm analysis overlap limit value, and when the algorithm analysis overlap value is large, When the algorithm analysis overlap value is equal to the algorithm analysis overlap limit value, no processing is performed. When the algorithm analysis overlap value is less than the algorithm analysis overlap limit value, the ideal analysis times are increased by one, and the ideal analysis times are marked as Cgdj. The algorithm analysis values of two similar algorithm recognition records in the same group are calculated by difference and the absolute value is taken to obtain the algorithm analysis fluctuation value. The algorithm analysis fluctuation limit value is set. When the algorithm analysis fluctuation value is greater than or equal to the algorithm analysis fluctuation limit value, the algorithm analysis fluctuation times are increased by one. When the algorithm analysis fluctuation value is less than the algorithm analysis fluctuation limit value, no processing is performed, and the algorithm analysis fluctuation times are marked as Ttrd. The formula is used. The algorithm recommended value Dse of this type of face recognition algorithm is obtained, where y1 is the algorithm analysis mean coefficient, y2 is the ideal analysis number coefficient, and y3 is the algorithm analysis fluctuation number coefficient.
[0017] Compared with the prior art, the present invention has the following beneficial effects:
[0018] By setting up a face recognition simulation module, a recognition algorithm simulation module and a recognition algorithm optimization module, multiple face recognition simulation environments can be regularly constructed. Various face recognition algorithms can be simulated and recognized through the face recognition simulation environment, and the recognition effects of each face recognition algorithm in each face recognition simulation environment can be three-dimensionally and comprehensively analyzed. The face recognition algorithm with the best recognition ability can be comprehensively selected for the smart access control to perform face recognition in the current time period, and the face recognition algorithm can be replaced regularly to ensure the recognition accuracy and efficiency of the smart access control in the park. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] Figure 1 This is a flowchart of the operation process of a park smart access control face recognition device;
[0020] Figure 2 A flowchart for obtaining algorithm profiling values;
[0021] Figure 3 This is a module block diagram of a campus smart access control face recognition device. DETAILED DESCRIPTION
[0022] Reference Figures 1 to 3A park smart access control face recognition device includes a face recognition simulation module, a recognition algorithm simulation module, and a recognition algorithm optimization module.
[0023] Face recognition simulation module: preset time interval g (g is the duration of the time interval). When the time node of time interval g is reached, the environmental data around the access control is collected (environmental data includes but is not limited to lighting data and background data), and i basic face photos are set synchronously (basic face photos are preset photos and are only used to build a face recognition simulation environment). Based on the environmental data and the basic face photos, i digital twin models are built (each digital twin model is built based on environmental data and a basic face photo), and a face recognition simulation environment is generated based on the digital twin model.
[0024] Recognition algorithm simulation module: In each face recognition simulation environment, various types of face recognition algorithms are set to simulate and recognize basic face photos in turn, and then generate face algorithm recognition records for various types of face recognition algorithms. The face algorithm recognition records include the face recognition simulation environment number (a total of i face recognition simulation environments, i = 1, 2, ..., i, each face recognition simulation environment corresponds to an independent number), face recognition algorithm type, and algorithm analysis value.
[0025] Recognition algorithm optimization module: After all face recognition simulation environments complete simulation recognition, all face algorithm recognition records are collected, and then the algorithm recommendation values of each face recognition algorithm type are obtained. The face recognition algorithm type with the largest algorithm recommendation value is marked as the preferred recognition algorithm, and the access control performs face recognition based on the preferred recognition algorithm.
[0026] By setting up a face recognition simulation module, a recognition algorithm simulation module and a recognition algorithm optimization module, multiple face recognition simulation environments can be regularly constructed. Various face recognition algorithms can be simulated and recognized through the face recognition simulation environment, and the recognition effects of each face recognition algorithm in each face recognition simulation environment can be three-dimensionally and comprehensively analyzed. The face recognition algorithm with the best recognition ability can be comprehensively selected for the smart access control to perform face recognition in the current time period, and the face recognition algorithm can be replaced regularly to ensure the recognition accuracy and efficiency of the smart access control in the park.
[0027] The algorithm recommendation value of each face recognition algorithm type is obtained in the following way: obtain the face recognition algorithm type of the face algorithm recognition record, mark the face algorithm recognition records of the same face recognition algorithm type as the same algorithm recognition record, sum and average the algorithm analysis values of all the same algorithm recognition records to obtain the algorithm analysis mean Lksz, group all the same algorithm recognition records into one group, sum the algorithm analysis values of two same algorithm recognition records in the same group to obtain the algorithm analysis overlap value, set the algorithm analysis overlap limit value, the algorithm analysis overlap limit value is the system setting threshold, the specific size can be adjusted according to actual needs, when the algorithm analysis overlap value is greater than or equal to When the algorithm analysis overlaps the limit value, no processing is done. When the algorithm analysis overlap value is less than the algorithm analysis overlap limit value, the ideal analysis times are increased once, and the ideal analysis times are marked as Cgdj. The algorithm analysis values of two similar algorithm recognition records in the same group are calculated by difference and the absolute value is taken to obtain the algorithm analysis fluctuation value. The algorithm analysis fluctuation limit value is set. The algorithm analysis fluctuation limit value is the system setting threshold value. The specific size can be adjusted according to actual needs. When the algorithm analysis fluctuation value is greater than or equal to the algorithm analysis fluctuation limit value, the algorithm analysis fluctuation times are increased once. When the algorithm analysis fluctuation value is less than the algorithm analysis fluctuation limit value, no processing is done, and the algorithm analysis fluctuation times are marked as Ttrd. Using the formula The algorithm recommended value Dse of this type of face recognition algorithm is obtained, where y1 is the algorithm analysis mean coefficient, y2 is the ideal analysis frequency coefficient, and y3 is the algorithm analysis fluctuation frequency coefficient. The value of y1 is 0.71, the value of y2 is 0.76, and the value of y3 is 0.93.
[0028] The algorithm analysis value of the face recognition algorithm is obtained in the following way: when the face recognition algorithm recognizes the basic face photo, each time a face feature point is recognized, a face recognition analysis record of the face feature point is generated. The face recognition analysis record includes the face feature point name (the face feature point name is the name corresponding to the face feature point. The face feature point is the key part of face recognition, including but not limited to eyes, nose, and mouth), face recognition analysis value, recognition time (the recognition time is the corresponding time when the face feature point is recognized), and the face recognition analysis value is obtained in the following way: A rectangular coordinate system is constructed with the analysis value as the Y-axis and the face recognition time (the face recognition time is the entire time for recognizing the basic face photo) as the X-axis, and the points of the corresponding face recognition analysis values are marked in the rectangular coordinate system according to the recognition time. The points of two adjacent recognition times in the rectangular coordinate system are connected to obtain the face recognition analysis curve, and the two end points of the face recognition analysis curve are perpendicular to the X-axis to obtain two start and end analysis line segments, and the total area of the figure composed of the face recognition analysis curve, the two start and end analysis line segments and the X-axis is marked as the algorithm analysis value.
[0029] The face recognition analysis value of the face recognition analysis record is obtained in the following manner: when a facial feature point is identified, the feature data identified by the facial feature point is marked as real-recognition feature data, the feature data of the facial feature point in the basic face photo is extracted and marked as basic feature data, the real-recognition feature data and the basic feature data are preprocessed (the preprocessing method includes but is not limited to formatting), the processed real-time feature data and the basic feature data are packaged into a feature data comparison set, the feature data comparison set is used as input data of the face recognition analysis model, and the face recognition analysis model outputs the face recognition analysis value.
[0030] The face recognition analysis model is obtained in the following manner: obtaining s feature data comparison sets, in which the real-time feature data and basic feature data in the feature data comparison set can be real data or virtual data, constructing a deep learning model, using the feature data comparison set as training data for the deep learning model, assigning a face recognition analysis value to each training data, dividing the training data into a training set and a validation set in a 1:1 ratio, training the training set and the validation set, and obtaining a face recognition analysis model. The face recognition analysis value is greater than zero, and the larger the face recognition analysis value, the greater the deviation between the real-time feature data and the basic feature data, and the smaller the face recognition analysis value, the smaller the deviation between the real-time feature data and the basic feature data.
[0031] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.
[0032] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.
[0033] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.
[0034] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.
[0035] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0036] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0037] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.
[0038] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A park smart access control face recognition device, characterized in that: It includes face recognition simulation module, recognition algorithm simulation module and recognition algorithm optimization module; The face recognition simulation module is used to periodically generate i face recognition simulation environments; The recognition algorithm simulation module sets each type of face recognition algorithm in each face recognition simulation environment to perform simulation recognition on the basic face photo in turn, thereby generating face algorithm recognition records of each type of face recognition algorithm; After the recognition algorithm optimization module completes the simulation recognition in all face recognition simulation environments, the access control performs face recognition based on the optimization recognition algorithm.
2. According to claim 1, a campus smart access control face recognition device is characterized in that: Periodically generate i face recognition simulation environments, specifically: preset a time interval g, when the time node of the time interval g is reached, collect the environmental data around the access control, synchronously set i basic face photos, build i digital twin models based on the environmental data and the basic face photos, and generate a face recognition simulation environment based on the digital twin model.
3. The park smart access control face recognition device according to claim 1 is characterized in that: The face algorithm recognition record includes the face recognition simulation environment number, face recognition algorithm type, and algorithm analysis value.
4. The park smart access control face recognition device according to claim 3 is characterized in that: The algorithm analysis value of the face recognition algorithm is obtained in the following way: during the process of the face recognition algorithm identifying the basic face photo, each time a face feature point is identified, a face recognition analysis record of the face feature point is generated, a rectangular coordinate system is constructed with the face recognition analysis value as the Y-axis and the face recognition time as the X-axis, the point of the corresponding face recognition analysis value is marked in the rectangular coordinate system according to the recognition time, the points of two adjacent recognition times in the rectangular coordinate system are connected to obtain a face recognition analysis curve, the two end points of the face recognition analysis curve are perpendicular to the X-axis to obtain two start and end analysis line segments, and the total area of the figure composed of the face recognition analysis curve, the two start and end analysis line segments and the X-axis is marked as the algorithm analysis value.
5. The park smart access control face recognition device according to claim 4 is characterized in that: Face recognition analysis records include face feature names, face recognition analysis values, and recognition time.
6. The park smart access control face recognition device according to claim 5 is characterized in that: The face recognition analysis value of the face recognition analysis record is obtained in the following manner: when a face feature point is identified, the feature data identified by the face feature point is marked as real-recognition feature data, the feature data of the face feature point in the basic face photo is extracted and marked as basic feature data, the real-recognition feature data and the basic feature data are pre-processed, the processed real-time feature data and the basic feature data are packaged into a feature data comparison set, the feature data comparison set is used as input data of the face recognition analysis model, and the face recognition analysis model outputs the face recognition analysis value.
7. The campus smart access control face recognition device according to claim 1 is characterized in that: Collect all face algorithm recognition records, and then obtain the algorithm recommendation value of each face recognition algorithm type, and mark the face recognition algorithm type with the largest algorithm recommendation value as the preferred recognition algorithm.
8. The park smart access control face recognition device according to claim 7 is characterized in that: The algorithm recommendation value of each face recognition algorithm type is obtained in the following way: obtain the face recognition algorithm type of the face algorithm recognition record, mark the face algorithm recognition records of the same face recognition algorithm type as the same algorithm recognition record, sum up the algorithm analysis values of all the same algorithm recognition records and take the average to obtain the algorithm analysis mean Lksz, group all the same algorithm recognition records into one group, sum up the algorithm analysis values of two same algorithm recognition records in the same group to obtain the algorithm analysis overlap value, set the algorithm analysis overlap limit value, and when the algorithm analysis overlap value is greater than or equal to When the algorithm analysis overlaps the limit value, no processing is performed. When the algorithm analysis overlap value is less than the algorithm analysis overlap limit value, the ideal analysis times are increased by one, and the ideal analysis times are marked as Cgdj. The algorithm analysis values of two similar algorithm recognition records in the same group are calculated by difference and the absolute value is taken to obtain the algorithm analysis fluctuation value. The algorithm analysis fluctuation limit value is set. When the algorithm analysis fluctuation value is greater than or equal to the algorithm analysis fluctuation limit value, the algorithm analysis fluctuation times are increased by one. When the algorithm analysis fluctuation value is less than the algorithm analysis fluctuation limit value, no processing is performed, and the algorithm analysis fluctuation times are marked as Ttrd. The formula is used. The algorithm recommended value Dse of this type of face recognition algorithm is obtained, where y1 is the algorithm analysis mean coefficient, y2 is the ideal analysis number coefficient, and y3 is the algorithm analysis fluctuation number coefficient.