CN113780156A - Face recognition method and system based on cloud edge architecture - Google Patents
Face recognition method and system based on cloud edge architecture Download PDFInfo
- Publication number
- CN113780156A CN113780156A CN202111048739.6A CN202111048739A CN113780156A CN 113780156 A CN113780156 A CN 113780156A CN 202111048739 A CN202111048739 A CN 202111048739A CN 113780156 A CN113780156 A CN 113780156A
- Authority
- CN
- China
- Prior art keywords
- face
- library
- cloud
- terminal
- edge
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 36
- 230000003993 interaction Effects 0.000 claims abstract description 11
- 238000000605 extraction Methods 0.000 claims abstract description 10
- 238000004891 communication Methods 0.000 claims abstract description 8
- 230000009471 action Effects 0.000 claims abstract description 7
- 238000001514 detection method Methods 0.000 claims description 33
- 238000012545 processing Methods 0.000 claims description 23
- 230000008569 process Effects 0.000 claims description 7
- 230000008520 organization Effects 0.000 claims description 3
- 230000008901 benefit Effects 0.000 abstract description 4
- 238000007726 management method Methods 0.000 description 15
- 238000004364 calculation method Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 6
- 230000005540 biological transmission Effects 0.000 description 5
- 230000007547 defect Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000007246 mechanism Effects 0.000 description 3
- 230000004044 response Effects 0.000 description 3
- 230000001680 brushing effect Effects 0.000 description 2
- 238000013523 data management Methods 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000002567 autonomic effect Effects 0.000 description 1
- 230000036760 body temperature Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000001815 facial effect Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/60—Software deployment
- G06F8/65—Updates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
- G06F9/5072—Grid computing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C9/00—Individual registration on entry or exit
- G07C9/30—Individual registration on entry or exit not involving the use of a pass
- G07C9/32—Individual registration on entry or exit not involving the use of a pass in combination with an identity check
- G07C9/37—Individual registration on entry or exit not involving the use of a pass in combination with an identity check using biometric data, e.g. fingerprints, iris scans or voice recognition
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C9/00—Individual registration on entry or exit
- G07C9/30—Individual registration on entry or exit not involving the use of a pass
- G07C9/38—Individual registration on entry or exit not involving the use of a pass with central registration
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2209/00—Indexing scheme relating to G06F9/00
- G06F2209/50—Indexing scheme relating to G06F9/50
- G06F2209/502—Proximity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Quality & Reliability (AREA)
- Signal Processing (AREA)
- Computer Security & Cryptography (AREA)
- Human Computer Interaction (AREA)
- Collating Specific Patterns (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a face recognition method and a face recognition system based on a cloud side end architecture, wherein the cloud side end architecture comprises a terminal, a side end and a cloud end which are sequentially in communication connection, the terminal and the side end are both arranged in the same local network segment, and the cloud end is arranged in a remote network segment; the method comprises the steps that feature extraction is carried out on face information pushed by a terminal at an edge, and then comparison is carried out on the face information and face feature data transmitted to the edge from a cloud end, and a comparison result is obtained; controlling the action of the terminal according to the comparison result; and the cloud end manages and stores the comparison result of the edge end and performs data interaction with the edge end. Compared with the prior art, the method has the advantages of reducing the pressure of network bandwidth, improving the face recognition efficiency and reliability and the like.
Description
Technical Field
The invention relates to the technical field of face recognition, in particular to a face recognition method and system based on a cloud edge architecture.
Background
Along with the development of face recognition technology, face recognition is increasingly integrated into our daily life. How can the face recognition technology be deployed safely and effectively in the face of the situations of multi-remote distribution, multi-network segment division and multi-system interaction? Most of the existing face recognition solutions issue face information to a terminal, and algorithm processing is performed after the terminal acquires the face information; or the face information acquired by the terminal is transmitted to the cloud end, and algorithm processing is carried out at the cloud end; and a cloud side end fusion technology is also used, the side end is used for processing the face data, and the acquired face information is returned to the cloud end for comparison. All the schemes can realize the function of face recognition, but the implementation has some defects to be promoted.
The following prior art currently exists:
the first prior art is as follows: the solution of comparing the face algorithm at the terminal is as follows: after the terminal collects the face information, face detection and face characteristic value extraction are carried out on the terminal, an algorithm is called to obtain a comparison result, then the result is pushed to the terminal, and data are stored and managed in the terminal.
The first prior art has the following defects: 1. the face information and the related data are stored in the terminal, so that the face information is leaked, and the information safety cannot be ensured. 2. Since the system is an offline environment, it is not easy to synchronize basic information from a system such as OA, for example, increase/decrease of a person feature value, customized management of a threshold value, and the like. 3, the performance of the terminal device with the ARM architecture is very limited, and once the number of faces is large, the comparison time is greatly increased, which results in poor experience.
The second prior art is: the solution of face algorithm comparison at the cloud end is as follows: after the terminal collects the face information, the face information is transmitted back to the cloud, and algorithm processing comparison and data management are carried out at the cloud.
The second prior art has the following defects: the data needs to be transmitted back to the background, consuming a lot of time in the transmission process. Meanwhile, algorithm processing and data management are both carried out in a cloud terminal, so that the processing speed is too low due to too high cloud terminal computing pressure, and the coupling degree is high.
The prior art is three: by using the cloud side-end fusion technology, the terminal collects face information, the side end performs face detection, face tracking and quality detection, then a target face image corresponding to each face image is sent to the cloud end, and face recognition is performed at the cloud end.
The third prior art has the following defects: and the image is returned to the cloud after the image processing is carried out by the edge terminal, compared with the technology II that the image is directly transmitted without being processed, the redundant data can be reduced, but the burden can still be brought to a transmission network in the transmission process. The face recognition is carried out at the cloud end, and the computing pressure can be brought to the cloud end.
Disclosure of Invention
The invention aims to overcome the defects in the prior art, and provides a face recognition method and a face recognition system based on a cloud edge end architecture, so as to solve the problem of safety of face information of a user, reduce the pressure of large message interaction of a face image on network bandwidth, and improve the face recognition efficiency and reliability.
The purpose of the invention can be realized by the following technical scheme:
a face recognition method based on a cloud side end architecture comprises a terminal, a side end and a cloud end which are sequentially in communication connection, wherein the terminal and the side end are both installed in the same local network segment, the cloud end is installed in a remote network segment, and the face recognition method comprises the following steps:
and a data processing step of the terminal: acquiring face information in real time through a terminal, performing quality detection and living body detection on the face information, and pushing the face information to an edge terminal if the quality detection and the living body detection both pass;
and (3) data processing steps of the side end: the method comprises the steps that an edge receives face feature information transmitted by a cloud, feature extraction is carried out according to the face information transmitted by a terminal, and comparison is carried out on the face feature information and face feature data transmitted to the edge from the cloud to obtain a comparison result; controlling the action of the terminal according to the comparison result, and pushing the comparison result to the cloud for management;
cloud data processing: and counting, managing and storing the data transmitted by the side end, performing data interaction with the side end, and extracting face characteristic information according to the face data.
Further, the face feature information is stored in a biological information model, the biological information model is divided into a plurality of biological library systems, and the corresponding biological library systems are selected by the side end to carry out face comparison;
each biological library system is provided with a false recognition library and a basic library, each object in the false recognition library and the basic library is correspondingly provided with a face recognition threshold, and the face recognition threshold of the same object in the false recognition library is larger than the face recognition threshold in the basic library;
and when the side end carries out face comparison, searching in the false recognition library and the basic library in sequence until corresponding personnel are found, and if the score obtained by carrying out face comparison in the biological library system is greater than the corresponding face recognition threshold value, judging and recognizing the corresponding personnel.
Further, the plurality of biological library systems are divided according to organization, region, and location.
Furthermore, the biological library system is also provided with a blacklist library for identifying blacklist personnel, and when the side end carries out face comparison, the side end searches in the blacklist library, the false identification library and the basic library in sequence until finding out corresponding personnel.
Furthermore, the terminals are distributed in each scene of each park, and according to the type of the park, the scene requirements and the personnel information, the face data and the face recognition threshold of the misidentification library and the basic library of the biological library system are configured in the corresponding terminals, so that the distributed face recognition for each park is realized.
Further, the terminal also performs deduplication processing when acquiring the face information, and the deduplication processing specifically includes: and uploading the face information once every other preset first time interval.
Further, the data processing step of the terminal further includes: and if the comparison result returned by the side end is that the face recognition is successful, the face recognition result is not processed repeatedly within a preset second time interval.
The invention also provides a face recognition system based on the cloud edge terminal architecture, which comprises a terminal, an edge terminal and a cloud end which are sequentially connected in a communication manner, wherein the terminal and the edge terminal are both arranged in the same local network segment, the cloud end is arranged in a remote network segment,
the terminal is configured to: acquiring face information in real time through a terminal, performing quality detection and living body detection on the face information, and pushing the face information to an edge terminal if the quality detection and the living body detection both pass;
the edge terminal is configured to: the method comprises the steps that an edge receives face feature information transmitted by a cloud, feature extraction is carried out according to the face information transmitted by a terminal, and comparison is carried out on the face feature information and face feature data transmitted to the edge from the cloud to obtain a comparison result; controlling the action of the terminal according to the comparison result, and pushing the comparison result to the cloud for management;
the cloud is configured to: and counting, managing and storing the data transmitted by the side end, and performing data interaction with the side end.
Further, the face feature information is stored in a biological information model, the biological information model is divided into a plurality of biological library systems, and the corresponding biological library systems are selected by the side end to carry out face comparison;
each biological library system is provided with a false recognition library and a basic library, each object in the false recognition library and the basic library is correspondingly provided with a face recognition threshold, and the face recognition threshold of the same object in the false recognition library is larger than the face recognition threshold in the basic library;
and when the side end carries out face comparison, searching in the false recognition library and the basic library in sequence until corresponding personnel are found, and if the score obtained by carrying out face comparison in the biological library system is greater than the corresponding face recognition threshold value, judging and recognizing the corresponding personnel.
Furthermore, the terminals are distributed in each scene of each park, and according to the type of the park, the scene requirements and the personnel information, the face data and the face recognition threshold of the misidentification library and the basic library of the biological library system are configured in the corresponding terminals, so that the distributed face recognition for each park is realized.
Compared with the prior art, the invention has the following advantages:
(1) the cloud edge terminal architecture is utilized, the face information and the temperature information are collected through the terminal, the face information is compared through the edge terminal, the data are managed and stored through the remote terminal, the terminal and the edge terminal in the same local network segment are mainly used for data calculation, calculation is preposed, the data are processed nearby, transmission is safer, and data processing is more timely; the data are processed by using a plurality of edge end nodes, so that the response speed is higher and the calculation efficiency is higher; and the data is uniformly managed at the cloud end by deploying the edge end in different places, so that the management is convenient, and the safety of the data is improved.
(2) The flexible deployment is realized in the face of a complex use scene by using a multi-park mode; in addition, a multi-biological-library model is used, so that the speed and the accuracy of biological identification are enhanced; in the actual use process, a threshold value can be dynamically set according to the identification condition, and the passing rate and the false identification rate are guaranteed.
(3) By using the cloud edge end architecture, the universal server is used at the edge end, and a special server is not required to be used; the edge terminals are deployed to various parks nearby and then are centralized in a cloud end for unified management, so that the speed of face recognition is greatly improved; the version release uses two mechanisms of self-service upgrade and remote release, so that the version release is more flexible; two duplicate removal mechanisms are used at the terminal and the background, so that repeated identification is avoided, and the efficiency is increased; on the basis of using the cloud edge end architecture, related functions are improved by combining problems encountered in practical use, the speed and the efficiency of face recognition are increased, the interaction between management and each system is facilitated, and the safety of face recognition information is improved.
Drawings
Fig. 1 is a schematic network architecture diagram of a face recognition system based on a cloud edge architecture according to an embodiment of the present invention;
fig. 2 is a schematic physical architecture diagram of a face recognition system based on a cloud-edge architecture according to an embodiment of the present invention;
fig. 3 is a schematic cloud edge configuration diagram of a face recognition system based on a cloud edge architecture according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating a biological information model provided in an embodiment of the present invention;
fig. 5 is a schematic diagram of a multi-campus mode according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Abbreviations and key term definitions:
face recognition:
a biological characteristic recognition method using human face characteristics as identification identity. The identity of the user is identified by extracting the facial image features of the user and comparing the features with the stored template features.
Cloud:
the system consists of servers, stores, processes, sorts and analyzes data which cannot be processed by edge calculation, and realizes the interactive work with multiple systems.
Side end:
on one side close to an object or a data source, an operation program is completed by utilizing an edge zone close to the data source, and a large amount of data does not need to be uploaded to a cloud. The method can be deployed in intelligent devices and computing nodes of different magnitudes, and provides safe, reliable, low-delay, low-cost, easily-expanded and weakly-dependent local computing service.
A terminal:
the equipment end, face identification entrance guard's equipment promptly realizes face acquisition, face detection function.
Example 1
The embodiment provides a face recognition method based on a cloud edge end architecture, wherein the cloud edge end architecture comprises a terminal, an edge end and a cloud end which are sequentially in communication connection, the terminal and the edge end are both installed in the same local network segment, the cloud end is installed in a remote network segment, and the face recognition method comprises the following steps:
and a data processing step of the terminal: acquiring face information in real time through a terminal, performing quality detection and living body detection on the face information, and pushing the face information to an edge terminal if the quality detection and the living body detection both pass; the face temperature information can be collected through the terminal and pushed to the side end for storage;
and (3) data processing steps of the side end: the method comprises the steps that an edge receives face feature information transmitted by a cloud, feature extraction is carried out according to the face information transmitted by a terminal, and comparison is carried out on the face feature information and face feature data transmitted to the edge from the cloud to obtain a comparison result; and controlling the action of the terminal according to the comparison result, and pushing the comparison result to the cloud for management.
As a preferred embodiment, the terminal further performs a deduplication process when acquiring the face information, where the deduplication process specifically is: the face information is uploaded once every other preset first time interval, and the waste of time and resources caused by repeated recognition is avoided.
As a preferred embodiment, if the comparison result returned by the side terminal is that the face recognition is successful, the face recognition result is not repeatedly processed within a preset second time interval.
Cloud data processing: and counting, managing and storing the data transmitted by the side end, performing data interaction with the side end, and extracting face characteristic information according to the face data.
By utilizing a cloud edge terminal framework, calculation is prepositioned and distributed in terminals and edge terminals in the same local network, so that data can be processed nearby, transmission is safer, and data processing is more timely; the data are processed by using a plurality of edge end nodes, so that the response speed is higher and the calculation efficiency is higher; the data are uniformly managed at the cloud end by deploying the edge end in different places, so that the management is convenient, and the safety of the data is improved; meanwhile, the snapshot technology is optimized, so that the snapshot speed is higher, the accuracy is higher, and the user experience is greatly improved.
The face feature information is stored in a biological information model, the biological information model is divided into a plurality of biological library systems according to mechanisms, regions, positions and the like, and the corresponding biological library systems are selected by the side end to compare faces;
each biological library system is provided with a false recognition library and a basic library, each object in the false recognition library and the basic library is correspondingly provided with a face recognition threshold, and the face recognition threshold of the same object in the false recognition library is larger than the face recognition threshold in the basic library;
and when the side end carries out face comparison, searching in the false recognition library and the basic library in sequence until a corresponding person is found, and if the score obtained by carrying out face comparison in the biological library system is greater than a corresponding face recognition threshold value, judging to recognize the corresponding person.
By using the biological information model, the accuracy of biological identification can be improved, and false identification can be avoided; meanwhile, the related information such as names, certificates and the like of the personnel can be stored in each biological information base, so that the personnel information can be conveniently and quickly acquired after the personnel are found.
Preferably, the biological library system is further provided with a blacklist library for identifying blacklist personnel, and when the side end performs face comparison, the blacklist library, the misidentification library and the basic library are sequentially searched until corresponding personnel are found.
Preferably, the terminals are distributed in each scene of each campus, and according to the type of the campus, the scene requirements and the personnel information, the face data and the face recognition threshold of the misidentification library and the basic library of the biological library system are configured in the corresponding terminals, so that the distributed face recognition for each campus is realized.
Example 2
The embodiment provides a face recognition system based on a cloud edge terminal architecture, which comprises a terminal, an edge terminal and a cloud end, wherein the terminal, the edge terminal and the cloud end are sequentially connected in a communication manner, the terminal and the edge terminal are both installed in the same local network segment, the cloud end is installed in a remote network segment,
the terminal is configured to: acquiring face information and temperature information in real time, performing quality detection on the face information, performing living body detection according to the temperature information, and pushing the face information to an edge terminal if the quality detection and the living body detection both pass;
the edge terminal is configured to: the method comprises the steps that an edge receives face feature information transmitted by a cloud, feature extraction is carried out according to the face information transmitted by a terminal, and comparison is carried out on the face feature information and face feature data transmitted to the edge from the cloud to obtain a comparison result; controlling the action of the terminal according to the comparison result, and pushing the comparison result to the cloud for management;
the cloud is configured to: and counting, managing and storing the data transmitted by the side end, and performing data interaction with the side end.
The biological information model is divided into a plurality of biological library systems, and the corresponding biological library systems are selected by the side end to carry out face comparison;
each biological library system is provided with a false recognition library and a basic library, each object in the false recognition library and the basic library is correspondingly provided with a face recognition threshold, and the face recognition threshold of the same object in the false recognition library is larger than the face recognition threshold in the basic library;
and when the side end carries out face comparison, searching in the false recognition library and the basic library in sequence until a corresponding person is found, and if the score obtained by carrying out face comparison in the biological library system is greater than a corresponding face recognition threshold value, judging to recognize the corresponding person.
The following is a detailed description.
1. Network architecture
As shown in fig. 1, the terminal device and the edge server are installed in the same network segment, the cloud server is installed in a remote network segment, and a firewall is installed between the edge server and the remote network, so that direct communication between different network segments is realized.
2. Physical architecture
As shown in fig. 2, the terminal is composed of a face brushing device and a temperature measuring device, the face brushing terminal captures face information, and the face information captured is synchronously pushed to the side end after duplication removal. And the temperature measuring equipment acquires the temperature and then asynchronously pushes the temperature to the side end.
The method comprises the steps that a plurality of edge servers are deployed nearby at the edge, an algorithm is called to extract human face features and compare human faces, the comparison result is transmitted to a terminal to be displayed to a user, and meanwhile the comparison result is pushed to the cloud. In addition, the side end is also responsible for receiving relevant information transmitted from the cloud end.
The cloud deploys a plurality of cloud servers, receives information transmitted from the side end, stores the information in the database, and interacts with other systems.
3. Terminal configuration
As shown in fig. 3, the terminal performs functions such as image acquisition, quality detection and living body detection, and synchronously pushes captured face information to the side end, asynchronously pushes identified body temperature information to the side end, and the side end performs feature extraction on the face information and synchronously receives face feature information pushed by the cloud for face comparison; this scheme combines actual conditions, has carried out a series of optimizations at the terminal to face identification machine:
A. adding the function of removing the weight: because the detection object is probably in the detection range for a long time, a large number of repeated detection objects appear, so that a human face can be identified only once within a certain time, and the waste of time and resources caused by repeated identification is avoided.
B. Setting a request time interval: and (4) starting snapshot after entering the snapshot area, and not repeatedly recognizing after the face recognition is successful within a certain time.
C. There are two version upgrading modes: autonomic version upgrades and remote version releases.
Self-service version upgrading: and upgrading the version on the face recognition equipment through manual operation.
Remote version release: and remotely controlling version upgrading of the terminal on the cloud management page.
D. And (4) newly adding threshold management, flexibly setting a threshold according to the identification result of the user in practical application, and simultaneously ensuring the passing rate and the false identification rate.
E. Reducing the dependence on the algorithm: because the characteristic values extracted by various algorithms are different, the original method of transmitting the characteristic values to the edge end is changed into transmitting pictures, so that the algorithms can be compatible, and the dependence on the algorithms is reduced.
4. Edge terminal configuration
And a plurality of edge ends are arranged nearby to process data, so that the data are distributed in a dot shape. And the side terminal calls an algorithm to carry out feature extraction and face comparison on the information transmitted by the terminal to obtain a comparison result. And transmitting the result to a terminal response to control the switch of the gate and the display information of the terminal equipment, and simultaneously pushing the result to a cloud terminal for management. In addition, the system can also receive information transmitted from the cloud end for processing.
5. Cloud configuration
Connecting a database, and creating and managing a table required by the system; data returned by the multiple ground sides are unified at the cloud end for statistics and management, and the data are stored in the inflow water meter after duplication removal; connecting each system to realize interaction with multiple systems; the management platform is provided for managing data in the system, and can control remote version release.
6. Biological information model for face comparison in side terminal
As shown in fig. 4, a biological information model is provided in the edge for comparing the human face, the biological information model is divided into a plurality of biological library systems, and the biological library systems are divided according to factors such as organization, region, location, and the like; a black name list library, a false recognition library and a basic library are arranged in the divided biological library system. When people are searched in the corresponding biological library system, the people can be sequentially searched from the blacklist library, the false recognition library and the basic library in sequence until the corresponding people are found.
Each biometric library has a corresponding threshold, and generally, the thresholds of the blacklist library and the misrecognized library are higher than those of the base library, so as to improve the accuracy of biometric identification.
TABLE 1 comparative example Using misidentification library
Table 1 shows how misidentification can be avoided using the misidentification library: when the object A is compared with the face, the score of the object B is higher than the score of the object A, the algorithm takes the highest score of which the face score is higher than the threshold value of the query object as a recognition result, and therefore if a false recognition library is not used, the object B is mistakenly recognized as the object A (85 is greater than 83 and is greater than the threshold value 80), and recognition errors are caused. At this time, the object B should be placed in the false recognition library, and the threshold value (87) of the false recognition library is set to be higher than the comparison score (85) of the object A and the object B, so that the object B is not recognized, and the object A can be correctly recognized.
By using the biological information model, the accuracy of biological identification can be improved, and false identification can be avoided; meanwhile, the related information such as names, certificates and the like of the personnel can be stored in each biological library system, so that the personnel information can be conveniently and quickly acquired after the personnel are found.
7. Multi-campus mode layout of terminals and frontends
As shown in fig. 5, a multi-park mode is used in actual use, that is, various types of parks can be set according to actual needs, and one person can belong to a plurality of parks. In different types of parks, different scene construction is used as requirements, and configuration is carried out in the terminal.
The multi-park mode has the advantages that the number of personnel is large, the situation that the use scene is complex can be rapidly integrated, the deployment is rapid, the management is convenient, and the multi-park mode is used in various demand scenes on a trial basis.
The foregoing detailed description of the preferred embodiments of the invention has been presented. It should be understood that numerous modifications and variations could be devised by those skilled in the art in light of the present teachings without departing from the inventive concepts. Therefore, the technical solutions available to those skilled in the art through logic analysis, reasoning and limited experiments based on the prior art according to the concept of the present invention should be within the scope of protection defined by the claims.
Claims (10)
1. The utility model provides a face identification method based on cloud limit end architecture, cloud limit end architecture includes terminal, limit end and the high in the clouds of communication connection in proper order, its characterized in that, terminal and limit end all install in same local network section, the high in the clouds is installed in long-range network section, face identification method includes following step:
and a data processing step of the terminal: acquiring face information in real time through a terminal, performing quality detection and living body detection on the face information, and pushing the face information to an edge terminal if the quality detection and the living body detection both pass;
and (3) data processing steps of the side end: the method comprises the steps that an edge receives face feature information transmitted by a cloud, feature extraction is carried out according to the face information transmitted by a terminal, and comparison is carried out on the face feature information and face feature data transmitted to the edge from the cloud to obtain a comparison result; controlling the action of the terminal according to the comparison result, and pushing the comparison result to the cloud for management;
cloud data processing: and counting, managing and storing the data transmitted by the side end, performing data interaction with the side end, and extracting face characteristic information according to the face data.
2. The face recognition method based on the cloud edge architecture as claimed in claim 1, wherein the face feature information is stored in a biological information model, the biological information model is divided into a plurality of biological library systems, and the edge selects the corresponding biological library system to perform face comparison;
each biological library system is provided with a false recognition library and a basic library, each object in the false recognition library and the basic library is correspondingly provided with a face recognition threshold, and the face recognition threshold of the same object in the false recognition library is larger than the face recognition threshold in the basic library;
and when the side end carries out face comparison, searching in the false recognition library and the basic library in sequence until corresponding personnel are found, and if the score obtained by carrying out face comparison in the biological library system is greater than the corresponding face recognition threshold value, judging and recognizing the corresponding personnel.
3. The method according to claim 2, wherein the plurality of biological library systems are divided according to organization, region and location.
4. The method according to claim 2, wherein the biological library system further comprises a blacklist library for identifying blacklist persons, and when the edge performs face comparison, the edge sequentially searches the blacklist library, the false recognition library and the basic library until corresponding persons are found.
5. The face recognition method based on the cloud edge architecture according to claim 2, wherein the terminals are distributed in each scene of each campus, and face data and face recognition thresholds of a misidentification library and a basic library of a biological library system are configured in the corresponding terminals according to the type of the campus, the scene requirements and the personnel information, so that distributed face recognition for each campus is realized.
6. The face recognition method based on the cloud edge architecture as claimed in claim 1, wherein the terminal further performs a deduplication process when acquiring the face information, and the deduplication process specifically includes: and uploading the face information once every other preset first time interval.
7. The face recognition method based on the cloud-edge architecture as claimed in claim 1, wherein the data processing step of the terminal further includes: and if the comparison result returned by the side end is that the face recognition is successful, the face recognition result is not processed repeatedly within a preset second time interval.
8. A face recognition system based on a cloud edge terminal architecture is characterized by comprising a terminal, an edge terminal and a cloud end which are sequentially connected in a communication manner, wherein the terminal and the edge terminal are both arranged in the same local network segment, the cloud end is arranged in a remote network segment,
the terminal is configured to: acquiring face information in real time through a terminal, performing quality detection and living body detection on the face information, and pushing the face information to an edge terminal if the quality detection and the living body detection both pass;
the edge terminal is configured to: the method comprises the steps that an edge receives face feature information transmitted by a cloud, feature extraction is carried out according to the face information transmitted by a terminal, and comparison is carried out on the face feature information and face feature data transmitted to the edge from the cloud to obtain a comparison result; controlling the action of the terminal according to the comparison result, and pushing the comparison result to the cloud for management;
the cloud is configured to: and counting, managing and storing the data transmitted by the side end, performing data interaction with the side end, and extracting face characteristic information according to the face data.
9. The face recognition system based on the cloud edge architecture as claimed in claim 8, wherein the face feature information is stored in a biological information model, the biological information model is divided into a plurality of biological library systems, and the edge selects the corresponding biological library system to perform face comparison;
each biological library system is provided with a false recognition library and a basic library, each object in the false recognition library and the basic library is correspondingly provided with a face recognition threshold, and the face recognition threshold of the same object in the false recognition library is larger than the face recognition threshold in the basic library;
and when the side end carries out face comparison, searching in the false recognition library and the basic library in sequence until corresponding personnel are found, and if the score obtained by carrying out face comparison in the biological library system is greater than the corresponding face recognition threshold value, judging and recognizing the corresponding personnel.
10. The face recognition system based on the cloud edge architecture according to claim 9, wherein the terminals are distributed in each scene of each campus, and face data and face recognition thresholds of the misidentification library and the basic library of the bio-library system are configured in the corresponding terminals according to the type of the campus, the scene requirements and the personnel information, so that distributed face recognition for each campus is realized.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111048739.6A CN113780156A (en) | 2021-09-08 | 2021-09-08 | Face recognition method and system based on cloud edge architecture |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111048739.6A CN113780156A (en) | 2021-09-08 | 2021-09-08 | Face recognition method and system based on cloud edge architecture |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113780156A true CN113780156A (en) | 2021-12-10 |
Family
ID=78841675
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111048739.6A Pending CN113780156A (en) | 2021-09-08 | 2021-09-08 | Face recognition method and system based on cloud edge architecture |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113780156A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170148241A1 (en) * | 2014-07-28 | 2017-05-25 | Dan Kerning | Security and Public Safety Application for a Mobile Device with Audio/Video Analytics and Access Control Authentication |
CN109117745A (en) * | 2018-07-23 | 2019-01-01 | 青岛理工大学 | Cloud face recognition and positioning method based on building information model |
CN109492536A (en) * | 2018-10-12 | 2019-03-19 | 大唐高鸿信息通信研究院(义乌)有限公司 | A kind of face identification method and system based on 5G framework |
CN111400047A (en) * | 2020-03-18 | 2020-07-10 | 青岛牛利智能科技有限公司 | Method for detecting and identifying human face from monitoring video stream through cloud edge cooperation |
CN112749626A (en) * | 2020-12-10 | 2021-05-04 | 同济大学 | DSP platform-oriented rapid face detection and recognition method |
-
2021
- 2021-09-08 CN CN202111048739.6A patent/CN113780156A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170148241A1 (en) * | 2014-07-28 | 2017-05-25 | Dan Kerning | Security and Public Safety Application for a Mobile Device with Audio/Video Analytics and Access Control Authentication |
CN109117745A (en) * | 2018-07-23 | 2019-01-01 | 青岛理工大学 | Cloud face recognition and positioning method based on building information model |
CN109492536A (en) * | 2018-10-12 | 2019-03-19 | 大唐高鸿信息通信研究院(义乌)有限公司 | A kind of face identification method and system based on 5G framework |
CN111400047A (en) * | 2020-03-18 | 2020-07-10 | 青岛牛利智能科技有限公司 | Method for detecting and identifying human face from monitoring video stream through cloud edge cooperation |
CN112749626A (en) * | 2020-12-10 | 2021-05-04 | 同济大学 | DSP platform-oriented rapid face detection and recognition method |
Non-Patent Citations (1)
Title |
---|
陆娇蓝等: "基于云计算的嵌入式人脸识别***建构与研究", 《计算机测量与控制》, vol. 24, no. 4, 25 April 2016 (2016-04-25), pages 146 - 148 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110096996B (en) | Biological information identification method, device, terminal, system and storage medium | |
CN202940921U (en) | Real-time monitoring system based on face identification | |
CN102542299A (en) | Face recognition method, device and mobile terminal capable of recognizing face | |
CN107257161A (en) | A kind of transformer station's disconnecting link remote control auxiliary check method and system based on state recognition algorithm | |
CN103996026A (en) | Fingerprint feature extraction method, device and system | |
CN103107886A (en) | Identification recognition system and method | |
CN102917207A (en) | Motion sequence based abnormal motion vision monitoring system | |
CN112507314B (en) | Client identity verification method, device, electronic equipment and storage medium | |
CN109034418A (en) | Operation field information transferring method and system | |
CN115171260A (en) | Intelligent access control system based on face recognition | |
CN115376063A (en) | Intelligent anti-error method based on cloud edge cooperation and image recognition | |
CN113627321A (en) | Image identification method and device based on artificial intelligence and computer equipment | |
CN113780156A (en) | Face recognition method and system based on cloud edge architecture | |
CN209821849U (en) | Owl face recognition early warning system | |
CN104239891A (en) | Remote intelligent evidence obtaining system | |
CN114863364B (en) | Security detection method and system based on intelligent video monitoring | |
CN111241930A (en) | Method and system for face recognition | |
CN112235589B (en) | Live network identification method, edge server, computer equipment and storage medium | |
CN115391596A (en) | Video archive generation method and device and storage medium | |
WO2021129491A1 (en) | Pedestrian search method, server, and storage medium | |
CN114283361A (en) | Method and apparatus for determining status information, storage medium, and electronic apparatus | |
CN113158931A (en) | Low-voltage centralized reading, operation and maintenance implementation method based on AI intelligent face recognition technology | |
CN113128262A (en) | Target identification method and device, storage medium and electronic device | |
CN113569808A (en) | Computer room face recognition dynamic management method, storage medium and system | |
CN113192250A (en) | Access control image processing method and device and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |