CN115966001A

CN115966001A - Recognition method, system, electronic device, storage medium, and program product

Info

Publication number: CN115966001A
Application number: CN202111189143.8A
Authority: CN
Inventors: 杜波
Original assignee: Beijing Kuangshi Technology Co Ltd; Beijing Megvii Technology Co Ltd
Current assignee: Beijing Kuangshi Technology Co Ltd; Beijing Megvii Technology Co Ltd
Priority date: 2021-10-12
Filing date: 2021-10-12
Publication date: 2023-04-14

Abstract

The disclosure relates to an identification method, system, electronic device, storage medium and program product. The method comprises the steps of obtaining an underlying library through first equipment, and determining first category underlying library characteristics and second category underlying library characteristics based on attributes of the underlying library characteristics in the underlying library. The first equipment sends the first category base library characteristics to the recognition device, and sends the second category base library characteristics to the cloud. And identifying the image to be identified based on the identification device and/or the cloud. Therefore, the technical problem that the number of the human face bottom library images loaded by the recognition device is limited can be solved.

Description

Recognition method, system, electronic device, storage medium, and program product

Technical Field

The present disclosure relates to the field of image processing, and in particular, to an identification method, system, electronic device, storage medium, and program product.

Background

With the development of science and technology, image recognition technology has been widely applied to traffic scenes such as passage, attendance checking and security protection. When the recognition device samples the image to be recognized, the recognition device is triggered to execute the image recognition process. In the identification device, the image to be identified is compared with the bottom library to realize the verification business requirement.

In the related art, before performing identification and authentication in the identification device, all the base libraries need to be loaded into the memory of the identification device, so that when the image to be identified is compared with the base libraries, the comparison result can be obtained quickly. However, the recognition device is limited in the size, system and use environment, the amount of the base library loaded by the recognition device is limited, and if the base library exceeding the upper limit is loaded, the recognition device cannot run the recognition program.

Disclosure of Invention

To overcome the problems in the related art, the present disclosure provides an identification method, system, electronic device, storage medium, and program product.

According to a first aspect of the embodiments of the present disclosure, there is provided an identification method applied to a first device, the identification method including:

acquiring a bottom library and attributes of the characteristics of the bottom library in the bottom library; determining a first category of base features and a second category of base features based on the attributes of the base features in the base, wherein the identification authentication passing frequency of the first category of base features is greater than or equal to a frequency threshold, and the identification authentication passing frequency of the second category of base features is less than the frequency threshold; and sending the first category base features to a recognition device, and sending the second category base features to a cloud.

In one embodiment, the method further comprises: receiving an attribute of the characteristics of the bottom library input by a user through a configuration interface, wherein the configuration interface is used for the user to configure the attribute of the characteristics of the bottom library as a first category characteristic attribute of the bottom library or a second category characteristic attribute of the bottom library.

In one embodiment, the method further comprises: and when the base library has the base library characteristics with changed attributes, updating the version number of the base library and recording the base library characteristics with changed attributes.

In one embodiment, the sending the first category-based library features to a recognition device and the second category-based library features to a cloud includes: if the version number of the base library is detected to be updated, determining the characteristics of the base library with changed attributes; and informing the identification device and/or the cloud of the characteristics of the bottom library with the changed attributes.

In one embodiment, the method further comprises: receiving a request sent by a recognition device and/or a cloud, wherein the request is used for determining the characteristics of a base library with changed attributes; acquiring the version number of the current base and the version number of the base in the identification device and/or the cloud; determining the characteristics of the base library with changed attributes based on the version number of the current base library and the version number of the base library in the identification device and/or the cloud; and informing the identification device and/or the cloud of the characteristics of the bottom library with the changed attributes.

In one embodiment, notifying the recognition device and/or the cloud of the underlying library feature with the changed attribute includes: if the attribute of the partial bottom library features is changed from the first category bottom library feature attribute to the second category bottom library feature attribute, sending an instruction for deleting the partial bottom library features to the identification device, and sending an instruction for adding the partial bottom library features to the cloud; and if the attribute of the partial feature base is changed from the second type base feature attribute to the first type base feature attribute, sending an instruction for adding the partial base feature to the identification device, and sending an instruction for deleting the partial base feature to the cloud.

In one embodiment, the method further comprises: configuring an individual identification mode of an identification device; or configuring a cloud independent identification mode; or configuring a recognition device and a cloud end combined recognition mode, wherein the combined recognition mode comprises recognition by the recognition device firstly and then cloud end recognition, or the recognition device and the cloud end are recognized simultaneously.

According to a second aspect of the embodiments of the present disclosure, there is provided an identification method applied to an identification device, the identification method including:

receiving a first category base library characteristic sent by first equipment, wherein the first category base library characteristic is a base library characteristic of which the identification authentication passing frequency is greater than or equal to a frequency threshold; storing the first category base features; and if the image to be recognized is detected, recognizing the image to be recognized based on the first category base library characteristics.

In one embodiment, the method further comprises: and responding to the fact that the base library features corresponding to the image to be recognized do not exist in the first category base library features, sending the image to be recognized to a cloud end, recognizing the image to be recognized by the cloud end based on second category base library features, wherein the recognition and authentication passing frequency of the second category base library features is smaller than a frequency threshold value.

In one embodiment, the method further comprises: sending the image to be recognized to a cloud end, and recognizing the image to be recognized by the cloud end based on second category base characteristics, wherein the recognition and authentication passing frequency of the second category base characteristics is less than a frequency threshold value; if the identification device is detected to be successfully identified or the cloud end is successfully identified, outputting an identification success result; and if the self identification failure of the identification device and the cloud identification failure are detected, outputting an identification failure result.

According to a third aspect of the embodiments of the present disclosure, there is provided an identification method applied to a cloud, the identification method including:

receiving second category base library characteristics sent by first equipment, wherein the second category base library characteristics are base library characteristics with identification authentication passing frequency smaller than a frequency threshold; saving the second category base features; receiving an image to be identified sent by an identification device; and identifying the image to be identified based on the second category base features.

According to a fourth aspect of the embodiments of the present disclosure, there is provided an identification system, including a first device, an identification apparatus, and a cloud end;

the first device is to: acquiring an underlying library and attributes of underlying library features in the underlying library, determining first category underlying library features and second category underlying library features based on the attributes of the underlying library features in the underlying library, wherein the passing frequency of identification and authentication of the first category underlying library features is greater than or equal to a frequency threshold, the passing frequency of identification and authentication of the second category underlying library features is less than the frequency threshold, and sending the first category underlying library features to an identification device and sending the second category underlying library features to a cloud end; the identification means is for: receiving first category base library characteristics sent by the first equipment, and storing the first category base library characteristics; identifying the detected image to be identified based on the first category base features; the cloud is used for: receiving second category base library characteristics sent by the first equipment, and storing the second category base library characteristics; and identifying the image to be identified received from the identification device based on the second category base library characteristic.

According to a fifth aspect of embodiments of the present disclosure, there is provided an electronic apparatus including:

a processor; a memory for storing processor-executable instructions; wherein the processor is configured to perform the identification method of any one of the embodiments of the first aspect; or configured to perform the identification method according to any of the embodiments of the second aspect, or configured to perform the identification method according to any of the embodiments of the third aspect.

According to a sixth aspect of the embodiments of the present disclosure, there is provided a storage medium, wherein the storage medium stores instructions that, when executed by a processor of a first device, enable the first device to execute the recognition method according to any one of the first aspect, or, when executed by a processor of a recognition apparatus, enable the recognition apparatus to execute the recognition method according to any one of the second aspect, or, when executed by a processor of a cloud, enable the cloud to execute the recognition method according to any one of the third aspect.

According to a seventh aspect of embodiments of the present disclosure, there is provided a computer program product comprising a computer program which, when executed by a processor, enables the processor to perform the identification method of any one of the first aspects, or enables the processor to perform the identification method of any one of the second aspects, or enables the processor to perform the identification method of any one of the third aspects.

The technical scheme provided by the embodiment of the disclosure can have the following beneficial effects: and acquiring the bottom library through the first equipment, and determining the first category bottom library characteristics and the second category bottom library characteristics based on the attributes of the bottom library characteristics in the bottom library. The first equipment sends the first category base library characteristics to the recognition device, and sends the second category base library characteristics to the cloud. And identifying the image to be identified based on the identification device and/or the cloud. Therefore, the technical problem that the capacity of the bottom library loaded by the identification device is limited can be solved.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and together with the description, serve to explain the principles of the disclosure.

Fig. 1 is a flowchart illustrating an identification method applied to a first device according to an exemplary embodiment.

FIG. 2 is a flow chart illustrating an identification method according to an example embodiment.

FIG. 3 is a block diagram illustrating a flow of face recognition according to an example embodiment.

FIG. 4 is a flowchart illustrating a detection of a change in an underlying library version number according to an example embodiment.

Fig. 5 is a block diagram illustrating a first device according to an example embodiment.

FIG. 6 is a block diagram of an electronic device according to an embodiment of the invention.

Detailed Description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the exemplary embodiments below are not intended to represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.

In recent years, technical research based on artificial intelligence, such as computer vision, deep learning, machine learning, image processing, and image recognition, has been advanced significantly. Artificial Intelligence (AI) is an emerging scientific technology for studying and developing theories, methods, techniques and application systems for simulating and extending human Intelligence. The artificial intelligence subject is a comprehensive subject and relates to various technical categories such as chips, big data, cloud computing, internet of things, distributed storage, deep learning, machine learning and neural networks. Computer vision is used as an important branch of artificial intelligence, specifically, a machine is used for identifying the world, and the computer vision technology generally comprises technologies such as face identification, living body detection, fingerprint identification and anti-counterfeiting verification, biological feature identification, face detection, pedestrian detection, target detection, pedestrian identification, image processing, image identification, image semantic understanding, image retrieval, character identification, video processing, video content identification, behavior identification, three-dimensional reconstruction, virtual reality, augmented reality, synchronous positioning and map construction (SLAM), computational photography, robot navigation and positioning and the like. With the research and progress of artificial intelligence technology, the technology is applied to various fields, such as security, city management, traffic management, building management, park management, face passage, face attendance, logistics management, warehouse management, robots, intelligent marketing, computational photography, mobile phone images, cloud services, smart homes, wearable equipment, unmanned driving, automatic driving, smart medical treatment, face payment, face unlocking, fingerprint unlocking, testimony verification, smart screens, smart televisions, cameras, mobile internet, live webcasts, beauty treatment, medical beauty treatment, intelligent temperature measurement and the like.

The identification device utilizes technologies such as image processing and image identification to carry out identification and authentication in different service scenes. The business scenarios may include, but are not limited to, schools, stations, businesses, and the like. In order to solve the technical problem that the capacity of the base supported by the identification device is limited, the embodiment of the disclosure provides an identification method. And dividing the characteristics of the bottom library in the bottom library into the characteristics of the bottom library with high passing frequency and the characteristics of the bottom library with low passing frequency during identification and authentication by using the background server. And sending the bottom library characteristics with high passing frequency to the recognition device, and sending the bottom library characteristics with low passing frequency to the cloud. And identifying and authenticating the detected image to be identified based on the base library characteristics with high passing frequency in the identification device and the base library characteristics with low passing frequency in the cloud. It can be seen that through the combination of the recognition device and the cloud, the loading quantity of the characteristics of the base library can be increased, and the requirements of any ultra-large base library characteristics can be met. It should be noted that the frequency of passing the identification authentication is determined with respect to a frequency threshold set according to actual needs.

In the following embodiments of the present disclosure, for convenience of description, the background management server is characterized by the first device. The base library can store a plurality of base library features, and the base library features can be specifically images, or features which can uniquely represent the images and are extracted from the images by adopting a neural network model.

It should be noted that the cloud in the embodiments of the present disclosure includes one or more servers, which are also referred to as a server cluster. The identification device and the server cluster have a corresponding relationship, and joint identification of the identification device and the server cluster can be realized based on the corresponding relationship. In order to further improve the identification efficiency, the identification device and the server cluster with the corresponding relation are deployed in the same local area network, so that the transmission time between the identification device and the server cluster can be saved, the identification efficiency is improved, the passing speed of a user is ensured, and the user has good passing experience. The identification device in the embodiment of the present disclosure may be a panel machine, an intelligent gate machine, a human evidence verification device, etc. The target object identified by the identification method provided by the embodiment of the disclosure can be a human face, a human body, a dog face, a palm print, a fingerprint, a building and other objects needing identification. It should be understood that the characteristics of the underlying library stored in the underlying library are correspondingly different according to the different recognition objects. Taking the identified target object as a face example, the features of the base library stored in the base library may be face images, or face features extracted from the face images.

The recognition device is exemplified by a panel machine, and in the related art, the panel machine can only support a 5-10W bottom library at most for face recognition. And the total base required by some scenes exceeds the limit of 10W, even reaches the level of 20W, 30W and the like, and if the panel machine completely loads the base exceeding the limit of 10W under the scene, the system is directly crashed, and the technical problem that the number of the bases supported by the panel machine is limited to about 10W and the identification of the oversized bases of 20W, 50W, 100W and the like cannot be met can be solved by using the identification method provided by the invention.

The following embodiments will explain the identification method provided by the present disclosure with reference to the drawings.

The embodiment of the disclosure provides an identification method, which is applied to first equipment. Fig. 1 is a flowchart illustrating an identification method applied to a first device according to an exemplary embodiment. The first device may be implemented as a backend server. As shown in fig. 1, the identification method applied to the first device includes the following steps S11 to S13.

In step S11, an underlying library, and attributes of the underlying library features in the underlying library are obtained.

In the embodiments of the present disclosure, the base library is provided according to an actual scene. In an embodiment of the present disclosure, attributes of the library features in the library are configured by a user through a configuration interface provided by the first device.

In one embodiment, the method comprises the steps of receiving an attribute of an underlying library feature input by a user through a configuration interface, wherein the configuration interface is used for enabling the user to configure the attribute of the underlying library feature as a first category underlying library feature attribute or a second category underlying library feature attribute. It should be noted that, for convenience of description, the first-category underlying library feature attribute is configured for the underlying library feature with the high frequency of passing the identification authentication. And configuring a second category bottom library feature attribute for the bottom library features with low passing frequency of identification authentication.

In step S12, a first category of the underlying library features and a second category of the underlying library features are determined based on the attributes of the underlying library features in the underlying library.

In the embodiment of the present disclosure, the passing frequency of the identification authentication of the first category-based library features is greater than or equal to the frequency threshold, and the passing frequency of the identification authentication of the second category-based library features is less than the frequency threshold. In other words, the bottom library features in the bottom library are classified according to the attributes of the bottom library features, and a first category bottom library feature and a second category bottom library feature are obtained.

For example, when the base library is acquired, an attribute configuration interface is provided, and the user configures the characteristic attribute of the base library according to the passing frequency of the identification authentication. And configuring the bottom library characteristic attribute with the identification authentication passing frequency larger than or equal to the frequency threshold as the first category bottom library characteristic attribute. And configuring the bottom library characteristic attribute with the identification authentication passing frequency smaller than the frequency threshold value as the second category bottom library characteristic attribute. The frequency threshold is set according to the actual scene needs, so that a user can flexibly adjust the quantity of the characteristics of the background library configured as the characteristic attribute of the first-class background library and the quantity of the characteristics of the background library configured as the characteristic attribute of the second-class background library.

In step S13, the first category base features are sent to the recognition device, and the second category base features are sent to the cloud.

And sending the first category base characteristics to an identification device so as to identify the image to be identified by utilizing the first category base characteristics through the identification device, thereby ensuring good passing experience. And sending the second category base features to a cloud end so as to identify the image to be identified by utilizing the second category base features through the cloud end, thereby ensuring the comprehensiveness of identification based on the base.

In one embodiment, to ensure real-time performance, the first device may send the first category-based library features to the identification device via MQTT (message queue telemetry transport) and/or send the second category-based library features to the cloud via MQTT.

The first equipment acquires the base library, marks the version number of the base library, and updates the version number of the base library if the attributes of the base library feature are increased, deleted and/or changed. The characteristics of the underlying library that cause changes to the version number of the underlying library are recorded. And transmitting the characteristics of the base library which cause the version number of the base library to change to the identification device and/or the cloud terminal in time.

And when the base library features with the changed attributes exist in the base library, updating the version number of the base library and recording the base library features with the changed attributes. It should be noted that the base library feature with changed attributes may be one or more.

In the embodiment of the present disclosure, if it is detected that the version number of the base is updated, the base feature with the changed attribute is determined, and the identification device and/or the base feature with the changed cloud attribute is notified. The reason for changing the attribute of the base library feature is not limited to a specific event, or the relationship between the passing frequency and the frequency threshold value of the identification authentication of the image to be identified corresponding to the base library feature is detected to change. For example, the identification method applied to the school is that in the graduation season of the school, the base library characteristics of graduates are changed from high pass frequency to low pass frequency, namely, the base library characteristic attributes of the graduates need to be changed from the first category base library characteristic attributes to the second category base library characteristic attributes. In another example, within a preset time interval, if the first device detects that the identification authentication pass frequency of one or part of the second category underlying library features is greater than the frequency threshold, the attribute of the underlying library feature is updated from the second category underlying library feature attribute to the first category underlying library feature attribute.

In order to avoid that the properties of the base library are not updated in time due to the fact that the identification device and/or the cloud end are disconnected from the first device. Therefore, the setting recognition device and/or the cloud sends a request to the first device at set time intervals.

In the embodiment of the disclosure, a first device receives a request sent by a recognition device and/or a cloud, and the request is used for determining a base library characteristic with a changed attribute; acquiring the version number of the current base and the version number of the base in the identification device and/or the cloud; determining the characteristics of the base library with changed attributes based on the version number of the current base library and the version number of the base library in the identification device and/or the cloud; and informing the identification device and/or the cloud end of the characteristics of the bottom library with changed attributes.

For example, when the first device sends the first category underlying library feature to the identification apparatus, the first device also sends the version number of the underlying library corresponding to the first category underlying library feature. When receiving a request sent by the identification device, the first device obtains the version number of the library in the current first device and the version number of the library in the identification device. And determining whether the version numbers of the base libraries are the same or not based on the version number of the base library in the current first equipment and the version number of the base library in the identification device. If the version numbers of the base libraries are the same, the identification device is informed that the base library characteristics with changed attributes do not exist. And if the version numbers of the base libraries are different, determining the base library characteristics of the base library characteristics with changed attributes. The underlying library features identifying the device attribute change are notified.

For example, in the embodiment of the present disclosure, when the first device sends the second-type base feature to the cloud, the first device also sends the version number of the base corresponding to the second-type base feature. When a request sent by a cloud is received, the first device obtains the version number of the library in the current first device and the version number of the library in the cloud. And determining whether the version numbers of the base libraries are the same or not based on the version number of the base library in the current first device and the version number of the base library in the cloud. If the version numbers of the base libraries are the same, the cloud is informed, and the characteristics of the base libraries with changed attributes do not exist. And if the version numbers of the base libraries are different, determining the characteristics of the base libraries with changed attributes. And informing the cloud end of the characteristics of the bottom library with changed attributes.

In an embodiment of the present disclosure, notifying the identifying device and/or the characteristics of the changed cloud end attribute of the base library includes: if the attribute of the partial base features is changed from the first category base feature attribute to the second category base feature attribute, sending an instruction for deleting the partial base features to the identification device, and sending an instruction for adding the partial base features to the cloud; and if the attribute of the partial feature base is changed from the second category base feature attribute to the first category base feature attribute, sending an instruction for adding the partial base feature to the identification device, and sending an instruction for deleting the partial base feature to the cloud. And continuing to send the change of the attribute of the bottom library characteristic of the graduate in the previous example, sending a command of deleting the bottom library characteristic of the graduate to the identification device, and sending a command of adding the bottom library characteristic of the graduate to the cloud. The situation that the characteristics of the bottom library of graduates are stored in the recognition device and the cloud end at the same time is avoided.

In one embodiment, the base library features correspond to unique identification numbers. For example, an instruction for deleting the first category base library features is sent to the recognition device, and the instruction comprises the identification number corresponding to the first category base library features to be deleted. And sending an instruction for deleting the second category base features to the cloud, wherein the instruction comprises an identification number corresponding to the second category base features to be deleted.

In the embodiment of the present disclosure, in addition to that the attribute of the characteristics of the base library is changed, which results in the update of the version number of the base library, there is any one of the following cases or a combination thereof that also results in the update of the version number of the base library:

the method comprises the steps of adding a first category base characteristic in a base library, or adding a second category base characteristic in the base library, or deleting the first category base characteristic in the base library, or deleting the second category base characteristic in the base library. Therefore, if the version number of the base is detected to be updated, the base characteristic causing the update of the version number of the base is determined. And if the first category base library features are newly added in the base library, sending the newly added first category base library features to the identification device. And if the base library features of the second category are newly added, sending the newly added base library features of the second category to the cloud. And if the first category base features are deleted from the base library, sending an instruction for deleting the first category base features to the identification device. And if the second type of base features are deleted from the base, sending an instruction for deleting the second type of base features to the cloud.

And the first equipment manages the characteristics of the bottom library in the bottom library by using the identification method. Namely, the input, the change and the cache of the first category base library characteristic and the second category base library characteristic are completed according to actual requirements, and the loading of the ultra-large base library is realized.

In the embodiment of the disclosure, the identification method also configures an individual identification mode of an identification device; or configuring a cloud independent identification mode; or, configuring a recognition device and a cloud joint recognition mode. The joint identification mode comprises identification by the identification device and then cloud identification, or the identification device and the cloud are identified at the same time.

The mode of individual identification by the identification device is suitable for the situation that the second category base library characteristics are all invalid or the user only concerns the first category base library characteristics. For example, in order to limit the flow of people, only the identification target passing right corresponding to the first category base library characteristic is granted. The cloud independent identification mode is suitable for the situation that the first category of underlying library features are all invalid or the user only cares about the second category of underlying library features. For example, it is known that the corresponding underlying library features of the image to be recognized exist in the scene of the underlying library features of the second category. And the identification mode is configured according to the scene requirement, so that the passing experience of the user can be improved.

Based on the same inventive concept, the invention also provides an identification method applied to the identification device, and the identification method comprises the following steps:

receiving a first category base library characteristic sent by first equipment, wherein the identification and authentication passing frequency of the first category base library characteristic is greater than or equal to a frequency threshold; storing the first category base library characteristics; and if the image to be recognized is detected, recognizing the image to be recognized based on the first-class bottom library characteristics. The recognition device performs recognition based on the trained recognition model. The recognition model is trained based on the first class base features.

In the embodiment of the disclosure, the identification mode configured by the first device is detected, and the image to be identified is identified according to the identification mode configured by the first device.

In one embodiment, when it is detected that the identification mode configured for the first device is the individual identification by the identification device, the identification device identifies the image to be identified based on the first category base features, and outputs the identification result.

And when the recognition mode configured by the first equipment is detected to be cloud independent recognition, the recognition device sends the detected image to be recognized to the cloud, and the cloud recognizes the image to be recognized based on the second category base characteristics. And the cloud identifies the image to be identified to obtain an identification result. And sending the recognition result to the recognition device, and outputting the recognition result by the recognition device.

And when the recognition mode configured by the first equipment is detected to be recognition by the recognition device firstly and then cloud recognition is detected, the recognition device recognizes the image to be recognized based on the first-class bottom library characteristics. And if the identification device identifies the image to be identified to obtain an identification success result, outputting the identification success result. If the identification device identifies the image to be identified to obtain an identification failure result, namely the first category of the bottom library features do not have the bottom library features corresponding to the image to be identified, the image to be identified is sent to the cloud end, and the cloud end identifies the image to be identified based on the second category of the bottom library features. And the cloud identifies the image to be identified to obtain an identification result. And sending the recognition result to the recognition device, and outputting the recognition result by the recognition device.

When the recognition mode configured by the first device is detected to be that the recognition device and the cloud terminal recognize at the same time, the image to be recognized is sent to the cloud terminal when the recognition device detects the image to be recognized. And identifying the image to be identified in the identification device based on the first category base library characteristics. And identifying the image to be identified on the basis of the second category base features at the cloud. If the recognition device is detected to be successfully recognized or the cloud recognition is successfully recognized, outputting a recognition success result; and if the self identification failure and the cloud identification failure of the identification device are detected, outputting an identification failure result.

In the embodiment of the disclosure, the face image to be recognized is recognized in the recognition device and the cloud. And if the recognition device or the cloud end obtains a successful recognition result, outputting the successful recognition result on the recognition device. And if the recognition device and the cloud end both obtain recognition failure results, outputting the recognition failure results on the recognition device. Under the condition of successful identification, the identification process is not required to be executed by the identification device and the cloud, and the identification of the face image to be identified can be stopped when the identification result is obtained, so that the identification efficiency can be improved, the overall passing experience can be ensured, and the computing power of the cloud or the identification device can be saved.

In the embodiment of the present disclosure, if the recognition device is provided with a display screen, the recognition result is displayed on the display screen. If the identification device is provided with a loudspeaker, the identification result is broadcasted through the loudspeaker. If the identification device is provided with a display screen and a loudspeaker, the identification result is displayed on the display screen and broadcasted through the loudspeaker.

Based on the same inventive concept, the embodiment of the present disclosure further provides an identification method, which is applied to a cloud, and includes: receiving second category base library characteristics sent by the first equipment, wherein the identification and authentication passing frequency of the second category base library characteristics is smaller than the frequency threshold; saving the second category base features; receiving an image to be identified sent by an identification device; and identifying the image to be identified based on the second category base characteristics. Therefore, the comprehensiveness of identifying the image to be identified is ensured, the problem that the characteristic of loading the base library by the identification device is limited is solved, and the service scene comprising the characteristics of the ultra-large base library is enabled.

In the embodiment of the present disclosure, the image to be recognized is an image recognized simultaneously with the recognition device, or the base library features corresponding to the image to be recognized are not stored in the first category base library features, and the first category base library features are sent to the recognition device by the first device.

Based on the same invention concept, the invention also provides an identification method, which is applied to identifying the image to be identified through the first equipment, the identification device and the cloud. FIG. 2 is a flow chart illustrating an identification method according to an example embodiment. As shown in fig. 2, the identification method includes the following steps.

In step S21, the first device obtains the base library and the attributes of the base library features in the base library, and determines the first category base library features and the second category base library features based on the attributes of the base library features in the base library.

In step S22, the first device sends the first category base features to the recognition device, and sends the second category base features to the cloud.

In step S23, the identification device receives and stores the first category-based library feature transmitted by the first device.

In step S24, the cloud receives and stores the second category library feature sent by the first device.

In step S25, the recognition device detects the recognition image, obtains a recognition mode configured by the first device, and if the first device is configured with a mode of joint recognition of the recognition device and the cloud, sends the image to be recognized to the cloud when the image to be recognized is recognized in the recognition device based on the first category base feature to obtain a recognition failure result.

In step S26, the cloud receives the image to be recognized sent by the recognition device, recognizes the recognition image based on the second category base features to obtain a recognition result, and sends the recognition result to the recognition device.

In step S27, the recognition device receives and outputs the recognition result.

In the embodiment of the present disclosure, the passing frequency of the identification authentication of the first category-based library feature is greater than or equal to the frequency threshold, and the passing frequency of the identification authentication of the second category-based library feature is less than the frequency threshold.

Taking face recognition as an example, fig. 3 is a block diagram illustrating a flow of face recognition according to an exemplary embodiment. As shown in fig. 3, the base library features are divided into first category base library features with high face authentication passing frequency and second category base library features with low face authentication passing frequency by using the first device. The first equipment sends the first category base features to the recognition device and sends the second category base features to the cloud. And carrying out face recognition based on the combination of the recognition device and the cloud.

In one embodiment, it is detected that the version number of the base is updated due to the change of the attribute of the base feature, and the first device notifies the panel machine and/or the cloud of the base feature with the changed attribute. FIG. 4 is a flowchart illustrating a detection of a change in an underlying library version number, according to an example embodiment. As shown in fig. 4, the following steps are included after detecting the change of the version number.

In step S31, the first device updates the version number of the base library and records the characteristics of the base library with changed attributes.

Take the change of the attribute of the base library feature a from the first class feature attribute to the second class feature attribute as an example.

In step S32, the first device sends an instruction to delete the base library feature a to the identification means.

In step S33, the first device sends an instruction to add the base library feature a to the cloud.

Based on the same invention concept, the invention also provides an identification system, which comprises first equipment, an identification device and a cloud end;

the first device is for: acquiring an underlying library and the attribute of the underlying library characteristic in the underlying library, determining a first category of underlying library characteristic and a second category of underlying library characteristic based on the attribute of the underlying library characteristic in the underlying library, wherein the passing frequency of identification and authentication of the first category of underlying library characteristic is greater than or equal to a frequency threshold, the passing frequency of identification and authentication of the second category of underlying library characteristic is less than the frequency threshold, transmitting the first category of underlying library characteristic to an identification device, and transmitting the second category of underlying library characteristic to a cloud end;

the identification means is for: receiving first-class base features sent by first equipment, and storing the first-class base features; identifying the detected image to be identified based on the first category base features;

the cloud is used for: receiving second category base characteristics sent by the first equipment, and storing the second category base characteristics; and identifying the image to be identified received from the identification device based on the second category base features.

Based on the same concept, the embodiment of the disclosure also provides a first device.

It is understood that the face recognition device provided by the embodiments of the present disclosure includes a hardware structure and/or a software module for performing the above functions. The disclosed embodiments can be implemented in hardware or a combination of hardware and computer software, in combination with the exemplary elements and algorithm steps disclosed in the disclosed embodiments. Whether a function is performed as hardware or computer software drives hardware depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.

Referring to fig. 5, fig. 5 is a block diagram illustrating a first device according to an example embodiment. The first device 100 comprises a first acquisition unit 101 and a processing unit 102.

A first obtaining unit 101, configured to obtain an underlying library and attributes of the underlying library features in the underlying library; the processing unit 102 is configured to determine a first category of library features and a second category of library features based on attributes of the library features in the library, send the first category of library features to the identification device, and send the second category of library features to the cloud, where the frequency of passing identification and authentication of the first category of library features is greater than or equal to a frequency threshold, and the frequency of passing identification and authentication of the second category of library features is less than the frequency threshold.

In one embodiment, the processing unit 102 is further configured to: and when the base library features with the changed attributes exist in the base library, updating the version number of the base library and recording the base library features with the changed attributes.

In one embodiment, the processing unit 102 is further configured to: if the version number of the base library is detected to be updated, determining the characteristics of the base library with changed attributes; and the base library characteristics inform the identification device and/or the cloud end attribute of change.

In one embodiment, the processing unit 102 is further configured to: receiving a request sent by an identification device and/or a cloud end, wherein the request is used for determining the characteristics of the bottom library with changed attributes; acquiring the version number of the current base and the version number of the base in the identification device and/or the cloud; determining the characteristics of the base library with changed attributes based on the version number of the current base library and the version number of the base library in the identification device and/or the cloud; and informing the identification device and/or the bottom library characteristics of the changed cloud end attributes.

Based on the same inventive concept, the invention also provides an identification device, which comprises a first receiving unit, a second receiving unit and a third receiving unit, wherein the first receiving unit is used for receiving the first category-based library characteristics sent by the first equipment, and the identification authentication passing frequency of the first category-based library characteristics is greater than or equal to a frequency threshold; the first storage unit is used for storing the first category base library characteristics; and the first identification unit is used for identifying the image to be identified based on the first category base characteristics if the image to be identified is detected.

Based on the same inventive concept, the invention also provides a cloud device, comprising:

the second receiving unit is used for receiving the second category base features sent by the first equipment, wherein the identification and authentication passing frequency of the second category base features is smaller than the frequency threshold; the second storage unit is used for storing the second category base library characteristics; the second identification unit is used for receiving the image to be identified sent by the identification device; and identifying the image to be identified based on the second category base features.

Based on the same inventive concept, the invention also provides an electronic device, comprising: a processor; a memory for storing processor-executable instructions; the processor is configured to execute a recognition method executed by the recognition device, or configured to execute a recognition method executed by the first device, or configured to execute a recognition method executed by the cloud. When the processor is configured to execute the method executed by the recognition device, the electronic equipment may further include a camera and a communication unit, wherein the camera is used for acquiring an image to be recognized; the communication unit is used for communicating with the associated equipment.

As shown in fig. 6, one embodiment of the present disclosure provides an electronic device 200. The electronic device 200 includes a memory 201, a processor 202, and an Input/Output (I/O) interface 203. The memory 201 is used for storing instructions. The processor 202 is configured to call the instruction stored in the memory 201 to perform the above-described recognition method performed by the first device, or the recognition method performed by the recognition apparatus, or the recognition method performed by the cloud. The processor 202 is connected to the memory 201 and the I/O interface 203, respectively, for example, via a bus system and/or other connection mechanism (not shown). The memory 201 may be used to store programs and data, including a program of the image data cleansing method or the face recognition method according to the embodiments of the present disclosure, and the processor 202 executes various functional applications and data processing of the electronic device 200 by running the program stored in the memory 201.

In the embodiment of the present disclosure, the processor 202 may be implemented in at least one hardware form of a Digital Signal Processor (DSP), a Field-Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), and the processor 202 may be one or a combination of a Central Processing Unit (CPU) or other Processing units with data Processing capability and/or instruction execution capability.

Memory 201 in the disclosed embodiments may comprise one or more computer program products that may include various forms of computer-readable storage media, such as volatile memory and/or non-volatile memory. The volatile Memory may include, for example, a Random Access Memory (RAM), a cache Memory (cache), and/or the like. The nonvolatile Memory may include, for example, a Read-Only Memory (ROM), a Flash Memory (Flash Memory), a Hard Disk Drive (HDD), a Solid-State Drive (SSD), or the like.

In the embodiment of the present disclosure, the I/O interface 203 may be used to receive input instructions (e.g., numeric or character information, and generate key signal inputs related to user settings and function control of the electronic apparatus 200, etc.), and may also output various information (e.g., images or sounds, etc.) to the outside. The I/O interface 203 in the disclosed embodiments may include one or more of a physical keyboard, function keys (such as volume control keys, switch keys, etc.), a mouse, a joystick, a trackball, a microphone, a speaker, a touch panel, and the like.

Another embodiment of the present disclosure also provides a computer program product, which includes a computer program, when the computer program is executed by a processor, the processor is enabled to execute the above-described identification method performed by the first device, or the identification method performed by the identification apparatus, or the identification method performed by the cloud.

It is to be understood that although operations are depicted in the drawings in a particular order, this is not to be understood as requiring that such operations be performed in the particular order shown or in serial order, or that all illustrated operations be performed, to achieve desirable results. In certain environments, multitasking and parallel processing may be advantageous.

The methods and apparatus related to embodiments of the present disclosure can be accomplished with standard programming techniques with rule-based logic or other logic to accomplish the various method steps. It should also be noted that the words "means" and "module," as used herein and in the claims, is intended to encompass implementations using one or more lines of software code, and/or hardware implementations, and/or equipment for receiving inputs.

Any of the steps, operations, or procedures described herein may be performed or implemented using one or more hardware or software modules, alone or in combination with other devices. In one embodiment, the software modules are implemented using a computer program product comprising a computer readable medium containing computer program code, which is executable by a computer processor for performing any or all of the described steps, operations, or procedures.

The foregoing description of the implementations of the disclosure has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the disclosure to the precise form disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of the disclosure. The embodiments were chosen and described in order to explain the principles of the disclosure and its practical application to enable one skilled in the art to utilize the disclosure in various embodiments and with various modifications as are suited to the particular use contemplated.

Claims

1. An identification method applied to a first device, the identification method comprising:

acquiring a bottom library and attributes of the characteristics of the bottom library in the bottom library;

determining a first category of base features and a second category of base features based on the attributes of the base features in the base, wherein the identification authentication passing frequency of the first category of base features is greater than or equal to a frequency threshold, and the identification authentication passing frequency of the second category of base features is less than the frequency threshold;

and sending the first category base features to a recognition device, and sending the second category base features to a cloud.

2. The identification method according to claim 1, characterized in that the method further comprises:

receiving an attribute of the characteristics of the bottom library input by a user through a configuration interface, wherein the configuration interface is used for the user to configure the attribute of the characteristics of the bottom library as a first category characteristic attribute of the bottom library or a second category characteristic attribute of the bottom library.

3. The identification method according to claim 1 or 2, characterized in that the method further comprises:

and when the base library has the base library characteristics with changed attributes, updating the version number of the base library and recording the base library characteristics with changed attributes.

4. The identification method according to claim 3, wherein the sending the first category-based library features to an identification device and the second category-based library features to a cloud comprises:

if the version number of the base library is detected to be updated, determining the characteristics of the base library with changed attributes;

and informing the identification device and/or the cloud of the characteristics of the bottom library with the changed attributes.

5. The identification method according to claim 3 or 4, characterized in that the method further comprises:

receiving a request sent by a recognition device and/or a cloud, wherein the request is used for determining the characteristics of a base library with changed attributes;

acquiring the version number of the current base and the version number of the base in the identification device and/or the cloud;

determining the characteristics of the base library with changed attributes based on the version number of the current base library and the version number of the base library in the identification device and/or the cloud;

6. The identification method according to claim 4 or 5, wherein notifying the identification device and/or the cloud of the underlying library feature with the changed attribute comprises:

if the attribute of the partial bottom library features is changed from the first category bottom library feature attribute to the second category bottom library feature attribute, sending an instruction for deleting the partial bottom library features to the identification device, and sending an instruction for adding the partial bottom library features to the cloud;

and if the attribute of the partial bottom library is changed from the second type bottom library feature attribute to the first type bottom library feature attribute, sending an instruction for adding the partial bottom library feature to the identification device, and sending an instruction for deleting the partial bottom library feature to the cloud.

7. The identification method according to claim 1, characterized in that the method further comprises:

configuring an individual identification mode of an identification device; or

Configuring a cloud independent identification mode; or

And configuring a recognition device and a cloud end joint recognition mode, wherein the joint recognition mode comprises recognition by the recognition device firstly and then cloud end recognition, or the recognition device and the cloud end are recognized simultaneously.

8. An identification method is applied to an identification device, and the identification method comprises the following steps:

receiving first category base library characteristics sent by first equipment, wherein the identification and authentication passing frequency of the first category base library characteristics is greater than or equal to a frequency threshold;

saving the first category base library characteristics;

and if the image to be recognized is detected, recognizing the image to be recognized based on the first-class bottom library features.

9. The identification method according to claim 8, characterized in that the method further comprises:

responding to the fact that the base library features corresponding to the images to be recognized do not exist in the first category base library features, sending the images to be recognized to a cloud end, recognizing the images to be recognized by the cloud end based on the second category base library features, wherein the recognition and authentication pass frequency of the second category base library features is smaller than a frequency threshold value.

10. The identification method according to claim 8, characterized in that the method further comprises:

sending an image to be recognized to a cloud end, and recognizing the image to be recognized by the cloud end based on second category base characteristics, wherein the passing frequency of recognition and authentication of the second category base characteristics is less than a frequency threshold value;

if the recognition device is detected to be successfully recognized or the cloud recognition is successfully recognized, outputting a recognition success result;

and if the self identification failure of the identification device and the cloud identification failure are detected, outputting an identification failure result.

11. The identification method is applied to a cloud end, and comprises the following steps:

receiving second category base library characteristics sent by first equipment, wherein the identification and authentication pass frequency of the second category base library characteristics is smaller than the frequency threshold;

saving the second category base features;

receiving an image to be identified sent by an identification device;

and identifying the image to be identified based on the second category base features.

12. An identification system, characterized in that the identification system comprises a first device, an identification apparatus and a cloud end;

the first device is to: acquiring an underlying library and attributes of underlying library features in the underlying library, determining first category underlying library features and second category underlying library features based on the attributes of the underlying library features in the underlying library, wherein the passing frequency of identification and authentication of the first category underlying library features is greater than or equal to a frequency threshold, the passing frequency of identification and authentication of the second category underlying library features is less than the frequency threshold, the first category underlying library features are sent to an identification device, and the second category underlying library features are sent to a cloud end;

the identification means is for: receiving first category base library characteristics sent by the first equipment, and storing the first category base library characteristics; identifying the detected image to be identified based on the first category base features;

the cloud is used for: receiving second category base library characteristics sent by the first equipment, and storing the second category base library characteristics; and identifying the image to be identified received from the identification device based on the second category base library characteristic.

13. An electronic device, comprising:

a processor;

a memory for storing processor-executable instructions;

wherein the processor is configured to perform the identification method of any one of claims 1 to 7; alternatively, the first and second electrodes may be,

is configured to perform the identification method of any one of claims 8 to 10, or,

is configured to perform the identification method of claim 11.

14. A storage medium having stored therein instructions that, when executed by a processor of a first device, enable the first device to perform the identification method of any one of claims 1 to 7, or,

the instructions in the storage medium, when executed by a processor of an identification apparatus, enable the identification apparatus to perform the identification method of any one of claims 8 to 10, or,

the instructions in the storage medium, when executed by a processor of a cloud, enable the cloud to perform the identification method of claim 11.

15. A computer program product, comprising a computer program which, when executed by a processor, enables the processor to carry out the identification method of any one of claims 1 to 7, or,

enabling a processor to perform the identification method of any one of claims 8 to 10, or,

enabling the processor to perform the identification method of claim 11.