Algorithm management method and system

Abstract

The application discloses an algorithm management method and system, which can respond to user configuration of an operation of a target algorithm with a target monitoring function, acquire an algorithm application package containing the target algorithm through an algorithm cloud platform, and acquire the target algorithm with the target monitoring function provided for the user from the algorithm cloud platform on the basis, so that the individualized monitoring demand of the user is met, only the target algorithm meeting the user demand is stored in the camera, and storage space is saved; the target algorithm can be pulled up from the algorithm application package through creation of a sub-process; subsequently, process communication can be established with the target algorithm, so that target monitoring information generated by the target algorithm and matched with the target monitoring function is acquired through the process communication. It can be seen that the application can save storage space, reduce hardware cost, and provide customized monitoring service for the user.

Description

Technical Field

[0001] The present application relates to the field of monitoring technology, and more specifically, to an algorithm management method and system. Background Art

[0002] In recent years, smart cameras have been widely used in the security industry. To meet the needs of different users, applications, and installation scenarios, commercially available smart cameras need to be compatible with a wide range of algorithm libraries. Existing technologies increase the storage capacity of smart cameras to accommodate as many algorithm libraries as possible. However, this solution can easily increase hardware costs, and users do not want redundant algorithm libraries in their smart cameras, which can cause lag.

[0003] Based on this, how to save storage space while meeting different needs has become the focus of people in this field. Summary of the Invention

[0004] In view of this, the present application provides an algorithm management method and system for saving storage space while meeting different monitoring requirements.

[0005] In order to achieve the above objectives, the following solutions are proposed:

[0006] An algorithm management method, applied to cloud-based middleware for cameras, includes:

[0007] In response to a user's operation of configuring a target algorithm with a target monitoring function, an algorithm application package containing the target algorithm is obtained through the algorithm cloud platform;

[0008] Pulling the target algorithm from the algorithm application package by creating a child process;

[0009] Establish process communication with the target algorithm to obtain target monitoring information generated by the target algorithm and matching the target monitoring function through process communication.

[0010] Optionally, obtaining an algorithm application package containing the target algorithm through an algorithm cloud platform includes:

[0011] Uploading a request instruction indicating the target monitoring function, the remaining resources of the camera, and the camera firmware version to the algorithm cloud platform;

[0012] Receive an algorithm application package issued by the algorithm cloud platform that matches the request instruction.

[0013] Optionally, also include:

[0014] Monitor the running data of the target algorithm in real time, and send heartbeat detection packets to the target algorithm at regular intervals to obtain the detection data of the target algorithm;

[0015] Based on the operation data and the detection data, determine whether the target algorithm has abnormal operation. If abnormal operation exists, upload the operation data, detection data, camera firmware version and target algorithm identification to the algorithm cloud platform accordingly.

[0016] Optionally, also include:

[0017] In response to the user's operation of setting the operating parameters, the operating parameters are transmitted to the target algorithm through process communication to adjust the target monitoring function of the target algorithm.

[0018] Optionally, also include:

[0019] All algorithms of the camera are prioritized regularly, and N algorithms with the lowest priorities are deleted.

[0020] Optionally, also include:

[0021] When the version of the first algorithm needs to be updated, determining a first identifier corresponding to the first algorithm;

[0022] Upload the update instruction containing the first identifier, the remaining resources of the camera and the camera firmware version to the algorithm cloud platform to obtain the latest version of the first algorithm matching the cloud middleware from the algorithm cloud platform.

[0023] Optionally, also include:

[0024] In response to a user operation of turning off a second monitoring function, determining a second algorithm matching the second monitoring function;

[0025] Determining whether the second algorithm is in an operating state;

[0026] If yes, a stop signal is issued and the second algorithm is closed;

[0027] If not, the second algorithm is disabled.

[0028] An algorithm management system, including a user terminal, an algorithm cloud platform, and cloud middleware in a camera;

[0029] The user terminal is used to request a target algorithm with a target monitoring function from the algorithm cloud platform;

[0030] The algorithm cloud platform is configured to, upon receiving a request from a user terminal, send an algorithm application package containing the target algorithm to the camera matched with the user terminal;

[0031] The cloud middleware is used to pull up the target algorithm from the algorithm application package by creating a sub-process, and establish process communication with the target algorithm, so as to obtain target monitoring information generated by the target algorithm and matching the target monitoring function through process communication.

[0032] Optionally, a monitoring cloud platform is also included;

[0033] The user terminal is further configured to request a target monitoring function from the monitoring cloud platform;

[0034] The monitoring cloud platform is used to generate a monitoring data scheduling instruction according to the target monitoring function, and send the monitoring data scheduling instruction to the cloud middleware;

[0035] The cloud middleware is further configured to determine whether there is target monitoring information that matches the monitoring data scheduling instruction. If not, the cloud middleware requests a target algorithm that matches the target monitoring function from the algorithm cloud platform, and after obtaining an algorithm application package containing the target algorithm, the cloud middleware creates a child process to pull up the target algorithm from the algorithm application package, establishes process communication with the target algorithm, and obtains the target monitoring information generated by the target algorithm and matching the monitoring data scheduling instruction through process communication; if so, the cloud middleware obtains the target monitoring information corresponding to the monitoring data scheduling instruction through process communication;

[0036] The algorithm cloud platform is further configured to respond to the cloud middleware and issue an algorithm application package that matches the cloud middleware and includes the target algorithm.

[0037] Optionally, the cloud middleware includes an algorithm request module:

[0038] The algorithm request module is used to upload a request instruction representing the target monitoring function, camera remaining resources and camera firmware version to the algorithm cloud platform, and receive an algorithm application package matching the request instruction issued by the algorithm cloud platform.

[0039] From the above technical solution, it can be seen that the algorithm management method provided by this application can respond to the user's operation of configuring a target algorithm with a target monitoring function, obtain an algorithm application package containing the target algorithm through the algorithm cloud platform, and on this basis, this application can obtain the target algorithm that provides the target monitoring function for the user from the algorithm cloud platform to meet the user's personalized monitoring needs. Only the target algorithm that meets the user's needs is stored in the camera, saving storage space; then, the target algorithm can be pulled up from the algorithm application package by creating a child process. On this basis, this application can run the target algorithm as an independent child process; then, process communication can be established with the target algorithm to obtain the target monitoring information generated by the target algorithm and matched with the target monitoring function through process communication; on this basis, the decoupling of each algorithm in the camera is achieved through cloud middleware, further reducing the difficulty of managing the camera algorithm. At the same time, an algorithm that meets user needs is provided, which can meet the needs of various users in various business scenarios while avoiding redundant algorithms. It can be seen that this application can provide users with customized monitoring services while saving storage space and reducing hardware costs, and can further reduce the difficulty of camera maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without any creative work.

[0041] Figure 1 A flow chart of an algorithm management method disclosed in an embodiment of this application;

[0042] Figure 2 This is a schematic diagram of the architecture of an algorithm management system disclosed in an embodiment of the present application;

[0043] Figure 3 This is a schematic diagram of the architecture of another algorithm management system disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0045] The algorithm management method of the present application can be applied to the cloud middleware of the camera, and the cloud middleware refers to a middleware software product that adapts to the cloud computing environment.

[0046] Next, combine Figure 1 The algorithm management method of this application is introduced in detail, including the following steps:

[0047] Step S1: In response to the user's operation of configuring a target algorithm with a target monitoring function, an algorithm application package containing the target algorithm is obtained through the algorithm cloud platform.

[0048] Specifically, when a user needs to configure a target monitoring function on a camera, an algorithm application package containing a target algorithm can be obtained through the algorithm cloud platform.

[0049] The user terminal can upload instructions indicating that the target monitoring function is configured on the camera or that the target monitoring function is viewed to the algorithm cloud platform, or can upload instructions indicating that the target monitoring function is configured on the camera or that the target monitoring function is viewed to the monitoring cloud platform. After the user terminal uploads the instruction, it can be determined that the user needs to configure the target monitoring function on the camera.

[0050] Step S2: Pull up the target algorithm from the algorithm application package by creating a subprocess.

[0051] Specifically, the target algorithm can be pulled from the algorithm application package and started by creating a subprocess to realize the target monitoring function.

[0052] Step S3: Establish process communication with the target algorithm to obtain target monitoring information generated by the target algorithm and matching the target monitoring function through process communication.

[0053] Specifically, a process communication can be established with the target algorithm, and through the process communication, the target monitoring information generated by the target algorithm can be retrieved and uploaded to the monitoring cloud platform for users to view.

[0054] Among them, process communication can be established by establishing a local socket.

[0055] It can be seen from the technical solution that the algorithm management method provided by the application can respond to the operation of the user configuring the target algorithm with the target monitoring function, acquire the algorithm application package containing the target algorithm through the algorithm cloud platform, and on this basis, the target algorithm providing the target monitoring function for the user can be acquired from the algorithm cloud platform to meet the personalized monitoring needs of the user, and only the target algorithm meeting the needs of the user is stored in the camera, thereby saving the storage space. Then, the target algorithm can be pulled up from the algorithm application package through the creation of a sub-process, and on this basis, the target algorithm can be run as an independent sub-process. Subsequently, process communication can be established with the target algorithm to acquire the target monitoring information generated by the target algorithm and matching the target monitoring function through the process communication. On this basis, the decoupling of each algorithm in the camera is realized through the cloud middleware, the difficulty of managing the camera algorithm is further reduced, and at the same time, the algorithm meeting the needs of the user is provided, which can meet the needs of various users in various business scenarios while avoiding redundant algorithms. It can be seen that the application can save storage space, reduce hardware cost, provide customized monitoring services for users, and further reduce the maintenance difficulty of the camera.

[0056] Through the cloud middleware, each algorithm in the camera is decoupled, each algorithm can be run in an independent process, mutual interference between each algorithm is reduced, and the running stability of the camera is improved.

[0057] In some embodiments of the application, the process of step S1, responding to the operation of the user configuring the target algorithm with the target monitoring function, acquiring the algorithm application package containing the target algorithm through the algorithm cloud platform, is described in detail, and the steps are as follows:

[0058] S10, uploading a request instruction indicating the target monitoring function, the remaining resources of the camera and the firmware version of the camera to the algorithm cloud platform.

[0059] Specifically, the cloud middleware can actively request the target algorithm, at this time, the request instruction indicating the target monitoring function, the remaining resources of the camera and the firmware version of the camera can be uploaded to the algorithm cloud platform.

[0060] The remaining resources of the camera can include the remaining memory of the camera, the remaining computing power of the camera, the remaining CPU of the camera and other remaining computing resources.

[0061] S11, receiving the algorithm application package matching the request instruction issued by the algorithm cloud platform.

[0062] Specifically, the algorithm application package returned by the algorithm cloud platform can be received, and the target algorithm matching the camera can be contained in the algorithm application package.

[0063] It can be seen from the above technical solution that this embodiment provides an optional way to obtain the algorithm application package. Through the above method, it is possible to actively request a target algorithm that conforms to the camera version and matches the user's needs, which can further meet the user's fragmented needs.

[0064] In some embodiments of the present application, the user's monitoring information is user privacy information. To ensure the security of the user's information, a process of monitoring the target algorithm can be added. Next, the process will be described in detail. The steps are as follows:

[0065] S4. Monitor the operating data of the target algorithm in real time, and periodically send a heartbeat detection packet to the target algorithm to obtain detection data of the target algorithm.

[0066] Specifically, the target algorithm in the camera can be monitored in real time, and the data generated by the target algorithm during operation can be used as operation data. At the same time, a scheduled task can be set to send the heartbeat detection package to the target algorithm through process communication at regular intervals to obtain the detection data generated by the target algorithm in response to the heartbeat detection package.

[0067] S5. Determine whether the target algorithm has abnormal operation based on the operation data and the detection data. If abnormal operation is found, upload the operation data, detection data, camera firmware version and target algorithm identifier to the algorithm cloud platform accordingly.

[0068] Specifically, it can be determined whether the target algorithm is operating abnormally based on the operating data and the detection data;

[0069] If there is abnormal operation, the operation data, detection data, camera firmware version and target algorithm identification will be uploaded to the algorithm cloud platform to retrieve the target algorithm for normal operation;

[0070] If there is no abnormal operation, the operation data will be reported to the algorithm cloud platform.

[0071] It can be seen from the above technical solutions that this embodiment newly adds an optional method for identifying whether the target algorithm is operating normally. Through the above method, it is possible to avoid user information security as much as possible while meeting user needs.

[0072] In some embodiments of the present application, considering that user needs and scene requirements may change, resulting in redundant algorithms in the camera, in order to improve the effective utilization of storage resources, a redundant algorithm deletion process can be added. The following is a detailed description of this process, and the steps are as follows:

[0073] S6. Regularly sort the priorities of all the algorithms of the camera and delete N algorithms with the lowest priorities.

[0074] Specifically, each algorithm may be prioritized according to its usage frequency, most recent usage time, and usage period, and the N algorithms with the lowest priority may be deleted.

[0075] N can be a non-negative integer.

[0076] In addition, the user can also actively delete the redundant algorithm. In response to the user's operation of deleting the redundant algorithm, the redundant algorithm in the running state can be stopped, the redundant algorithm can be deleted, and the process communication with the second algorithm can be deleted.

[0077] Among them, the user can send an instruction to delete redundant algorithms to the monitoring cloud platform, indicating that the redundant algorithms need to be deleted.

[0078] You can also use the LRU algorithm to remove redundant algorithms in the camera.

[0079] It can be seen from the above technical solution that this embodiment provides an optional method for deleting redundant algorithms. Through the above method, the utilization rate of the storage area can be improved by deleting redundant algorithms.

[0080] In some embodiments of the present application, considering that the camera algorithm may need to be updated to better implement the monitoring function, based on this, the present application adds a process for updating the algorithm. Next, the process will be described in detail. The steps are as follows:

[0081] S7. When the version of the first algorithm needs to be updated, determine a first identifier corresponding to the first algorithm.

[0082] Specifically, the first algorithm is an algorithm that needs to be updated in the camera.

[0083] The identification of the algorithm that needs to be updated can be determined.

[0084] S8. Upload the update instruction containing the first identifier, the remaining resources of the camera, and the camera firmware version to the algorithm cloud platform to obtain the latest version of the first algorithm that matches the cloud middleware from the algorithm cloud platform.

[0085] Specifically, the identifier of the algorithm that needs to be updated, the remaining resources of the camera, and the camera firmware version can be reported to the algorithm cloud platform to obtain the latest algorithm issued by the algorithm cloud platform that matches the firmware version.

[0086] It can be seen from the above technical solution that this embodiment provides an optional method for updating the algorithm. Through the above method, the monitoring function of the camera can be further improved to meet diverse monitoring needs as much as possible.

[0087] In some embodiments of the present application, considering that the user may actively turn off the monitoring function that is not currently needed, on this basis, in order to meet user needs, the present application adds a process for turning off the algorithm corresponding to the monitoring function. Next, the process will be described in detail. The steps are as follows:

[0088] S9. In response to the user's operation of turning off the second monitoring function, determine a second algorithm that matches the second monitoring function.

[0089] Specifically, the second monitoring function is a monitoring function that needs to be disabled, and the algorithm that matches the second monitoring function is the second algorithm.

[0090] S10. Determine whether the second algorithm is in operation. If so, execute step S11; if not, execute step S12.

[0091] Specifically, it may be determined whether the state of the second algorithm is a running state, if so, executing step S11, if not, executing step S12.

[0092] S11: Send a stop signal and close the second algorithm.

[0093] Specifically, when the second algorithm is in the running state, a stop signal may be sent to the second algorithm through process communication, and when the second algorithm stops running, the second algorithm is closed.

[0094] S12: Close the second algorithm.

[0095] Specifically, when the second algorithm is in the stopped state, the second algorithm can be directly turned off.

[0096] After closing the second algorithm, the user may need to open the closed algorithm during subsequent use. Therefore, the present application can also respond to the user's instruction to restart the third monitoring function by creating a sub-process to re-start the third algorithm that matches the third monitoring function.

[0097] The third monitoring function may be a monitoring function that needs to be restarted, and the third algorithm may be an algorithm that matches the third monitoring function.

[0098] It can be seen from the above technical solution that this embodiment provides an optional way to turn off the second monitoring function. Through the above method, the user's monitoring needs at different time periods can be further met, and customized monitoring services can be provided to users.

[0099] Next, we will combine Figure 2 The algorithm management system provided in this application is introduced in detail. The algorithm management system provided below can be compared with the algorithm management method provided above.

[0100] See also Figure 2 It can be found that the algorithm management system can include a user terminal 1, an algorithm cloud platform 2 and a cloud middleware 3.

[0101] The user terminal 1 may be a mobile terminal or a desktop computer used by the user.

[0102] The algorithm cloud platform 2 is a service platform set up on the network side, which can include algorithms with multiple different monitoring functions. The same monitoring function can also correspond to multiple versions of algorithms.

[0103] In addition, each algorithm can also correspond to the applicable camera model, cloud middleware model, version and algorithm description information.

[0104] The user terminal 1 can interact with the algorithm cloud platform 2 for data, and the algorithm cloud platform 2 can interact with the cloud middleware 3 for data.

[0105] Among them, the algorithm cloud platform 2 and the cloud middleware 3 can interact with each other through the signaling channel, and the cloud middleware 3 can monitor in real time through the local port whether there is interactive data in the signaling channel.

[0106] The data uploaded by the cloud middleware 3 to the algorithm cloud platform 2 may include algorithm-related information such as the cloud middleware model, camera ID, camera model, remaining camera resources, the function and version of each algorithm included in the camera, etc. The cloud middleware 3 may periodically report algorithm-related information.

[0107] The cloud middleware 3 can be loaded into the camera and belongs to the firmware of the camera.

[0108] The user terminal 1 can be used to request a target algorithm with target monitoring function from the algorithm cloud platform.

[0109] The algorithm cloud platform 2 can be used to send an algorithm application package containing the target algorithm to the camera matched with the user terminal after receiving a request from the user terminal.

[0110] Among them, the target algorithm can be an algorithm that matches the target monitoring function, the version of the cloud middleware 3, the camera model and the remaining resources of the camera.

[0111] The cloud middleware 3 can be used to pull up the target algorithm from the algorithm application package by creating a sub-process and establish process communication with the target algorithm to obtain the target monitoring information generated by the target algorithm and matching the target monitoring function through process communication.

[0112] The operation process of the target algorithm may require the configuration of operation parameters. The user terminal 1 can set the operation parameters when receiving the parameter configuration instructions issued by the algorithm cloud platform 2. The algorithm cloud platform 2 can send the operation parameters to the cloud middleware 3 to set the parameters of the target algorithm.

[0113] After the parameter setting is completed, the cloud middleware 3 can generate the operating data of the target algorithm and return it to the algorithm cloud platform 2. The algorithm cloud platform 2 can determine whether the target algorithm is running normally based on the operating data to determine whether the parameter setting is correct, and return the result to the user terminal 1.

[0114] Target monitoring information can be presented in the form of audio, alarms, snapshots, cloud playback videos and real-time videos.

[0115] The cloud middleware 3 can also obtain target monitoring information generated by the test run of the target algorithm and upload the target monitoring information to the monitoring cloud platform 4 so that users can view the parameter configuration results.

[0116] In addition to requesting the target algorithm of the target monitoring function from the algorithm cloud platform 2, the user terminal 1 can also request the monitoring cloud platform to configure the target monitoring function in the camera.

[0117] Therefore, the algorithm management system of the present application may also include a monitoring cloud platform.

[0118] This application provides another optional system architecture for implementing algorithm management, such as Figure 3 shown.

[0119] See also Figure 3 It can be found that the algorithm management system of the present application can also include a user terminal 1, an algorithm cloud platform 2, a cloud middleware 3 and a monitoring cloud platform 4.

[0120] The monitoring cloud platform 4 may be a service cloud platform configured on the network side.

[0121] The user terminal 1 can interact with the monitoring cloud platform 4 for data, and the monitoring cloud platform 4 can interact with the cloud middleware 3 for data.

[0122] The cloud middleware 3 can upload the monitoring information of each algorithm to the monitoring cloud platform 4 in real time.

[0123] The user terminal 1 can view the monitoring information through the monitoring cloud platform 4 in real time.

[0124] On this basis, the user terminal 1 can also be used to request the target monitoring function from the monitoring cloud platform;

[0125] The monitoring cloud platform 4 can be used to generate monitoring data scheduling instructions according to the target monitoring function, and send the monitoring data scheduling instructions to the cloud middleware;

[0126] The cloud middleware 3 can also be used to determine whether there is target monitoring information that matches the monitoring data scheduling instruction. If not, it requests the target algorithm that matches the target monitoring function from the algorithm cloud platform, and after obtaining the algorithm application package containing the target algorithm, it pulls up the target algorithm from the algorithm application package by creating a child process, establishes process communication with the target algorithm, and obtains the target monitoring information generated by the target algorithm and matching the monitoring data scheduling instruction through process communication; if so, it obtains the target monitoring information corresponding to the monitoring data scheduling instruction through process communication;

[0127] The algorithm cloud platform 2 is further configured to respond to the cloud middleware 3 and issue an algorithm application package that matches the cloud middleware and includes the target algorithm.

[0128] It can be seen from the above technical solutions that this embodiment provides two optional ways to update the camera algorithm. Through the above methods, the pertinence of this application is improved, providing users with a convenient and targeted monitoring function management solution.

[0129] Furthermore, the cloud middleware 3 may include an algorithm request module:

[0130] The algorithm request module is used to upload a request instruction representing the target monitoring function, camera remaining resources and camera firmware version to the algorithm cloud platform, and receive an algorithm application package matching the request instruction issued by the algorithm cloud platform.

[0131] Furthermore, the cloud middleware 3 may also include:

[0132] A detection data acquisition unit is used to monitor the operating data of the target algorithm in real time and send a heartbeat detection packet to the target algorithm at regular intervals to obtain the detection data of the target algorithm;

[0133] The abnormal operation judgment unit is used to determine whether the target algorithm has abnormal operation based on the operation data and the detection data. If abnormal operation exists, the operation data, detection data, camera firmware version and target algorithm identification are uploaded to the algorithm cloud platform accordingly.

[0134] Furthermore, the cloud middleware 3 may also include:

[0135] The parameter setting unit is used to respond to the user's operation of setting the operating parameters and transmit the operating parameters to the target algorithm through process communication to adjust the target monitoring function of the target algorithm.

[0136] Furthermore, the cloud middleware 3 may also include:

[0137] The algorithm deletion unit is used to regularly sort the priorities of all the algorithms of the camera and delete the N algorithms with the lowest priorities.

[0138] Furthermore, the cloud middleware 3 may also include:

[0139] an identification determining unit, configured to determine a first identification corresponding to the first algorithm when a version of the first algorithm needs to be updated;

[0140] An algorithm updating unit is used to upload an update instruction including the first identifier, remaining resources of the camera, and the camera firmware version to the algorithm cloud platform, so as to obtain the latest version of the first algorithm matching the cloud middleware from the algorithm cloud platform.

[0141] Furthermore, the cloud middleware 3 may also include:

[0142] an algorithm determining unit, configured to determine a second algorithm matching the second monitoring function in response to a user operation of turning off the second monitoring function;

[0143] The algorithm shut-down unit is configured to determine whether the second algorithm is in a running state; if so, send a stop signal and shut down the second algorithm; if not, shut down the second algorithm.

[0144] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0145] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0146] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. The various embodiments of the present application may be combined with each other. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An algorithm management method, characterized in that: Cloud middleware applied to cameras, which is a middleware software product installed in cameras and adapted to the cloud computing environment, includes: In response to a user's operation of configuring a target algorithm with a target monitoring function, an algorithm application package containing the target algorithm is obtained through the algorithm cloud platform; Pulling the target algorithm from the algorithm application package by creating a child process; Establishing process communication with the target algorithm to obtain target monitoring information generated by the target algorithm and matching the target monitoring function through the process communication; The method further comprises: In response to the user's operation of setting the operating parameters, the operating parameters are transmitted to the target algorithm through process communication to adjust the target monitoring function of the target algorithm.

2. The algorithm management method according to claim 1, characterized in that: The algorithm application package containing the target algorithm is obtained through the algorithm cloud platform, including: Uploading a request instruction indicating the target monitoring function, the remaining resources of the camera, and the camera firmware version to the algorithm cloud platform; Receive an algorithm application package issued by the algorithm cloud platform that matches the request instruction.

3. The algorithm management method according to claim 1 or 2, characterized in that: Also includes: Monitor the running data of the target algorithm in real time, and send heartbeat detection packets to the target algorithm at regular intervals to obtain the detection data of the target algorithm; Based on the operation data and the detection data, determine whether the target algorithm has abnormal operation. If abnormal operation exists, upload the operation data, detection data, camera firmware version and target algorithm identification to the algorithm cloud platform accordingly.

4. The algorithm management method according to claim 1 or 2, characterized in that: Also includes: All algorithms of the camera are prioritized regularly, and N algorithms with the lowest priorities are deleted.

5. The algorithm management method according to claim 1 or 2, characterized in that: Also includes: When the version of the first algorithm needs to be updated, determining a first identifier corresponding to the first algorithm; Upload the update instruction containing the first identifier, the remaining resources of the camera and the camera firmware version to the algorithm cloud platform to obtain the latest version of the first algorithm matching the cloud middleware from the algorithm cloud platform.

6. The algorithm management method according to claim 1 or 2, characterized in that: Also includes: In response to a user operation of turning off a second monitoring function, determining a second algorithm matching the second monitoring function; Determining whether the second algorithm is in an operating state; If yes, a stop signal is issued and the second algorithm is closed; If not, the second algorithm is disabled.

7. An algorithm management system, characterized in that: It includes user terminals, algorithm cloud platforms and cloud middleware in cameras. The cloud middleware is a middleware software product installed in cameras and adapted to the cloud computing environment. The user terminal is used to request a target algorithm with a target monitoring function from the algorithm cloud platform; The algorithm cloud platform is configured to, upon receiving a request from the user terminal, send an algorithm application package containing the target algorithm to the cloud middleware matched with the user terminal; The cloud middleware is used to pull the target algorithm from the algorithm application package by creating a child process and establish process communication with the target algorithm to obtain target monitoring information generated by the target algorithm and matching the target monitoring function through process communication; The cloud middleware further includes: The parameter setting unit is used to respond to the user's operation of setting the operating parameters and transmit the operating parameters to the target algorithm through process communication to adjust the target monitoring function of the target algorithm.

8. The algorithm management system according to claim 7, characterized in that: It also includes monitoring cloud platforms; The user terminal is further configured to request a target monitoring function from the monitoring cloud platform; The monitoring cloud platform is used to generate a monitoring data scheduling instruction according to the target monitoring function, and send the monitoring data scheduling instruction to the cloud middleware; The cloud middleware is further configured to determine whether there is target monitoring information that matches the monitoring data scheduling instruction. If not, the cloud middleware requests a target algorithm that matches the target monitoring function from the algorithm cloud platform, and after obtaining an algorithm application package containing the target algorithm, the cloud middleware creates a child process to pull up the target algorithm from the algorithm application package, establishes process communication with the target algorithm, and obtains the target monitoring information generated by the target algorithm and matching the monitoring data scheduling instruction through process communication. If so, obtaining target monitoring information corresponding to the monitoring data scheduling instruction through process communication; The algorithm cloud platform is further configured to respond to the cloud middleware and issue an algorithm application package that matches the cloud middleware and includes the target algorithm.

9. The algorithm management system according to claim 8, characterized in that: The cloud middleware includes the algorithm request module: The algorithm request module is used to upload a request instruction representing the target monitoring function, camera remaining resources and camera firmware version to the algorithm cloud platform, and receive an algorithm application package matching the request instruction issued by the algorithm cloud platform.

Images