WO2022135539A1 - 设备配置参数处理方法和装置、数据分析方法和装置、计算设备、计算机可读存储介质、以及计算机程序产品 - Google Patents
设备配置参数处理方法和装置、数据分析方法和装置、计算设备、计算机可读存储介质、以及计算机程序产品 Download PDFInfo
- Publication number
- WO2022135539A1 WO2022135539A1 PCT/CN2021/140929 CN2021140929W WO2022135539A1 WO 2022135539 A1 WO2022135539 A1 WO 2022135539A1 CN 2021140929 W CN2021140929 W CN 2021140929W WO 2022135539 A1 WO2022135539 A1 WO 2022135539A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- configuration parameter
- content analysis
- collection
- collection device
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 100
- 238000012545 processing Methods 0.000 title claims abstract description 60
- 238000003860 storage Methods 0.000 title claims abstract description 24
- 238000004590 computer program Methods 0.000 title claims abstract description 19
- 238000007405 data analysis Methods 0.000 title claims abstract description 17
- 238000004458 analytical method Methods 0.000 claims abstract description 200
- 230000004044 response Effects 0.000 claims abstract description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 48
- 238000013145 classification model Methods 0.000 claims description 48
- 238000012549 training Methods 0.000 claims description 47
- 230000008859 change Effects 0.000 claims description 42
- 230000002776 aggregation Effects 0.000 claims description 24
- 238000004220 aggregation Methods 0.000 claims description 24
- 238000006243 chemical reaction Methods 0.000 claims description 10
- 238000003672 processing method Methods 0.000 claims description 7
- 238000013480 data collection Methods 0.000 claims description 4
- 238000001514 detection method Methods 0.000 description 70
- 230000007547 defect Effects 0.000 description 48
- 238000010586 diagram Methods 0.000 description 23
- 230000003287 optical effect Effects 0.000 description 15
- 230000008569 process Effects 0.000 description 15
- 230000006870 function Effects 0.000 description 13
- 238000004519 manufacturing process Methods 0.000 description 13
- 238000007689 inspection Methods 0.000 description 8
- 238000013473 artificial intelligence Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 7
- 230000002950 deficient Effects 0.000 description 5
- 230000003993 interaction Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 230000004931 aggregating effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000007257 malfunction Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0803—Configuration setting
- H04L41/0813—Configuration setting characterised by the conditions triggering a change of settings
- H04L41/0816—Configuration setting characterised by the conditions triggering a change of settings the condition being an adaptation, e.g. in response to network events
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
- G05B19/41875—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by quality surveillance of production
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/02—Details
- H04L12/16—Arrangements for providing special services to substations
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0803—Configuration setting
- H04L41/0806—Configuration setting for initial configuration or provisioning, e.g. plug-and-play
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0893—Assignment of logical groups to network elements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/16—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence
Definitions
- the present disclosure generally relates to the technical field of data processing based on artificial intelligence, and in particular, to a method and apparatus for processing equipment configuration parameters, a method and apparatus for data analysis, a computing device, a computer-readable storage medium, and a computer program product.
- artificial intelligence analyzes the data provided by the Internet of Things devices through deep learning and other technologies, so as to realize the mining of data value.
- the application of IoT and artificial intelligence technology in the manufacturing industry can greatly improve the production efficiency of the manufacturing industry. Considering the particularity of the manufacturing process, it may be necessary to collect a large number of monitoring images for each manufacturing process for quality inspection of the process.
- the acquisition equipment of surveillance images and the artificial intelligence algorithm model for intelligent analysis of surveillance images are the key factors to ensure the quality inspection results. Changes in these key factors will lead to lower detection efficiency.
- a method for processing device configuration parameters includes: acquiring collection data of a collection device located at a target location, wherein the collection data is configured to be input into a content analysis model for the target location to obtain a content analysis result of the collected data; based on the collection data, to determine whether the current configuration parameters of the collection device have changed; in response to the change of the current configuration parameters of the collection device, send the preset configuration parameters for the target location to the collection device to be based on the The preset configuration parameter adjusts the current configuration parameter of the collection device; wherein, in the case where the input of the content analysis model is the collection data of the collection device configured with the preset configuration parameter, the content analysis model
- the output content analysis result has a first content analysis accuracy rate, and in the case where the input of the content analysis model is the collection data of the collection device configured with the current configuration parameter, the content analysis result output by the content analysis model There is a second content analysis accuracy rate, wherein the first content analysis accuracy rate is higher than the second content analysis accuracy
- determining whether the current configuration parameters of the collection device have changed includes: comparing the collected data with the same data indicators of historically collected data to obtain a first comparison result, wherein the The historical collection data is the data collected at the historical moment for the target location, and the historical moment is earlier than the data collection moment in time, and the data index is the change of the collected data with the configuration parameters of the collection device. and, based on the first comparison result, determining whether the current configuration parameter of the collection device has changed.
- determining whether the current configuration parameter of the collection device has changed based on the collected data includes: comparing the collected data with the same data indicators of the training data of the content analysis model to obtain a second comparison As a result, wherein the training data is data used to train the content analysis model, and the data index is an index of the collected data that changes with changes in configuration parameters of the collecting device; and, based on the second Comparing the results, it is determined whether the current configuration parameters of the collection device have changed.
- the collected data includes a current device identification of the collection device, and, based on the collected data, determining whether the current configuration parameter of the collection device has changed includes: from the collected data, identifying and determining whether the current configuration parameters of the collecting device have changed based on the device identification of the collecting device and the historical device identification of the target location.
- the method further includes: after acquiring the acquisition data of the acquisition device located at the target position, performing data format conversion on the acquisition data.
- determining whether the current configuration parameter of the collection device changes based on the collected data includes: inputting the collected data into a configuration parameter classification model to obtain the current configuration parameter category of the collection device; Determine whether the current configuration parameter category of the collection device is a preset configuration parameter category; in response to the current configuration parameter category of the collection device not being the preset configuration parameter category, determine that the current configuration parameter of the collection device has changed.
- sending the preset configuration parameters for the target position to the collection device includes: sending the preset configuration parameters corresponding to the preset configuration parameter categories to the collection device.
- the method further includes: acquiring first historical collection data for the target location and a configuration parameter category of a collection device that collects the first historical collection data; and, using the first historical collection data The data is used as a sample, and model training is performed with the configuration parameter category of the collection device that collects the first historical collection data as a label, so as to obtain the configuration parameter classification model.
- the method further includes: after acquiring the configuration parameter category of the collection device that collected the first historical collection data, inputting the configuration parameter category of the collection device that collected the first historical collection data into the configuration A parameter category aggregation model to obtain an aggregated configuration parameter category; and, using the first historically collected data as a sample and using the configuration parameter category of the collection device that collected the first historically collected data as a label to perform model training includes: Model training is performed using the first historically collected data as a sample and the aggregated configuration parameter category as a label to obtain the configuration parameter classification model.
- the method further includes: acquiring second historically collected data for the target location, a configuration parameter category of a collection device that collects the second historically collected data, and a description of the second historically collected data
- the target content result, the target content result is the objective content reflected by the second historical collection data
- the second historical collection data corresponding to each configuration parameter category is input into the content analysis model to obtain each The content analysis result of the second historically collected data corresponding to the configuration parameter category; based on the content analysis result of the second historically collected data corresponding to each configuration parameter category and the target content result, determine the content analysis model for each configuration parameter category.
- the content analysis accuracy rate of the second historically collected data of the configuration parameter category; the configuration parameter category corresponding to the highest content analysis accuracy rate is determined as the preset configuration parameter category.
- the method further includes: obtaining third historically collected data for the target location and a target content result of the third historically collected data, the target content result being the third historically collected data The objective content reflected; taking the third historically collected data as a sample, and using the target content result of the third historically collected data as a label to perform model training to obtain the content analysis model.
- the method is performed by an edge device, the edge device includes a rule engine service component and a device remote control service component, and, based on the collected data, it is determined whether the current configuration parameters of the collection device have changed Executed by the rule engine service component, and sending preset configuration parameters for the target location to the collection device is executed by the device remote control service component.
- the method is performed by an edge device and a platform server, and, based on the collected data, determining whether a current configuration parameter of the collection device has changed is performed by the edge device, and will be performed for the The preset configuration parameters of the target position are sent to the acquisition device for execution by the platform server.
- a data analysis method includes: acquiring the acquisition data of the acquisition device located at the target position; inputting the acquisition data into a configuration parameter classification model to obtain the configuration parameter category of the acquisition device; The content analysis model corresponding to the configuration parameter category of the device is used to obtain the content analysis result of the collected data.
- the method further includes: acquiring first historical collection data for the target location and a configuration parameter category of a collection device that collects the first historical collection data; taking the first historical collection data as sample, and perform model training with the configuration parameter category of the collection device that collects the first historical collection data as a label, so as to obtain the configuration parameter classification model.
- the method further includes: acquiring second historically collected data for the target location, a configuration parameter category of a collection device that collects the second historically collected data, and a description of the second historically collected data
- Target content result the target content result is the objective content reflected by the second historically collected data; for each configuration parameter category, the second historically collected data is taken as a sample, and the second historically collected data is taken as a sample.
- the target content result is a label for model training to obtain the content analysis model corresponding to each configuration parameter category.
- the collected data is input into a configuration parameter classification model to obtain a configuration parameter category of the collection device and the collected data is input into a content analysis model corresponding to the configuration parameter category of the collection device , so that the content analysis result of the collected data is executed by an algorithm model component, wherein the algorithm model component is arranged in an edge device or a platform server.
- a device configuration parameter processing apparatus includes an acquisition data acquisition module configured to acquire acquisition data from acquisition devices located at a target location, wherein the acquired data is configured to be input into a content analysis model for the target location to obtain the acquired data
- a configuration parameter change determination module which is configured to determine whether the current configuration parameters of the collection device have changed based on the collection data
- a preset configuration parameter sending module which is configured to respond to the collection device.
- the preset configuration parameters for the target position are sent to the collection device, so as to adjust the current configuration parameters of the collection device based on the preset configuration parameters; wherein, in the content
- the content analysis result output by the content analysis model has the first content analysis accuracy.
- the content analysis result output by the content analysis model has a second content analysis accuracy, wherein the first content analysis accuracy is higher than the first content analysis accuracy. 2. Content analysis accuracy.
- a data analysis device configured to include: an acquisition data acquisition module configured to acquire acquisition data of acquisition equipment located at a target location; a configuration parameter category determination module configured to input the acquired data into a configuration parameter classification model to obtain the acquired data A configuration parameter category of the device; and a content analysis module configured to input the collected data into a content analysis model corresponding to the configuration parameter category of the collection device to obtain a content analysis result of the collected data.
- a computing device comprising: a memory configured to store computer-executable instructions; a processor configured to when the computer-executable instructions are executed by the processor , to implement the method according to any embodiment of the present application.
- a computer-readable storage medium having computer-executable instructions stored thereon.
- the method according to any embodiment of the present application is implemented. .
- a computer program product comprising computer-executable instructions, wherein the computer-executable instructions, when executed by a processor, perform the method according to any of the embodiments of the present application.
- FIG. 1 shows a schematic diagram of an application scenario of a method for processing device configuration parameters provided by an embodiment of the present application
- FIG. 2 shows a schematic diagram of a system architecture of an edge device provided by an embodiment of the present application
- FIG. 3 shows an exemplary flowchart of a method for processing a device configuration parameter provided by an embodiment of the present application
- FIG. 4 shows an exemplary flowchart of a method for determining a preset configuration parameter provided by an embodiment of the present application
- FIG. 5 shows an exemplary flowchart of the data analysis method provided by the embodiment of the present application
- FIG. 6 shows a schematic diagram of an interaction flow of a method for processing a device configuration parameter provided by an embodiment of the present application
- FIG. 7 shows another schematic diagram of an interaction flow of a method for processing a device configuration parameter provided by an embodiment of the present application
- FIG. 8 shows another schematic diagram of an interaction flow of a method for processing a device configuration parameter provided by an embodiment of the present application
- FIG. 9 shows another schematic diagram of an interaction flow of a method for processing device configuration parameters provided by an embodiment of the present application.
- FIG. 10 is a schematic diagram of another interaction flow of the device configuration parameter processing method provided by the embodiment of the present application.
- FIG. 11 shows a schematic structural diagram of a device configuration parameter processing apparatus provided by an embodiment of the present application.
- FIG. 12 shows a schematic structural diagram of a data analysis apparatus provided by an embodiment of the present application.
- FIG. 13 shows a schematic structural diagram of a computing device provided by an embodiment of the present application.
- FIG. 14 shows a schematic structural diagram of a computer system of a device or a server provided by an embodiment of the present application.
- FIG. 1 shows an exemplary application scenario in which the method provided by the embodiment of the present application may be implemented.
- at least one collection device 101 at least one edge device 102 , and at least one platform server 103 are included.
- the collection device 101 may be configured to collect data, and the data collected by the collection device may be referred to as collection data.
- the collection device 101 includes, but is not limited to, a terminal device, a sensing device, and the like.
- the sensing device can be, for example, a camera, which can be applied in the industrial field, and specifically can detect a certain process step, for example, take an image related to the process step.
- the camera may photograph the product obtained after a certain process is completed to obtain an image.
- the image itself can objectively reflect the actual conditions of these products, such as whether these products are defective, and these conditions can also be referred to as the actual results reflected by these images.
- the sensing device may be an automatic optical inspection device (Automated Optical Inspection, AOI device for short).
- AOI equipment can detect common defects encountered in the production process based on optical principles.
- the sensing device may also be a temperature sensor, a humidity sensor, a pressure measuring device, etc., for collecting sensing data corresponding to a certain process step.
- Terminal devices include but are not limited to smart phones, tablet computers, televisions, notebook computers, desktop computers, virtual reality devices, etc. This application does not specifically limit this.
- Edge devices are used to provide physical connectivity and enable communication between networks, such as connecting an internal local area network (LAN) to the Internet (Internet) or an external wide area network (WAN).
- Traditional edge devices include edge routers, routing switches, firewalls, multiplexers, and other wide area network (WAN) devices.
- Intelligent edge devices have built-in processors with on-board analytics or artificial intelligence capabilities. By processing a certain amount of data directly on intelligent edge devices, rather than uploading, processing, and storing data on the cloud, efficiency can be improved and costs reduced.
- Edge devices 102 include, but are not limited to, edge gateways and edge servers.
- An edge gateway is a gateway deployed at the edge of a network. It connects the physical and digital worlds through functions such as network connection and protocol conversion, and provides lightweight connection management, real-time data analysis, and application management functions.
- edge devices 102 include edge servers. Edge servers provide users with a channel into the network and the ability to communicate with other server devices.
- an edge server can be a group of servers that perform a single function, such as a firewall server, a cache server, a load balancing server, a DNS server, and so on.
- FIG. 2 schematically shows the system architecture of the edge device 102 .
- the edge device 102 may adopt an open source edge computing framework (edge X Foundry).
- the open source edge computing framework can be designed in a hierarchical structure, from bottom to top, device services (which may include device service components), core services (which may include core service components), and support services (which may include rule service components and algorithms) model components) and application services.
- Open source edge computing frameworks can also include security services and management services (including device remote control service components).
- a device service layer can be a collection of concrete microservices that communicate directly with physical devices. Each device microservice can manage multiple physical devices that support the corresponding interface.
- the core service layer including services such as core-data, core-command, core-metadata, and registry&config.
- the support service layer including providing common services such as logs, rule engines, and reminders.
- the platform server 103 can be used to store the algorithm model, directly receive the collection data uploaded by the collection device, and analyze the uploaded collection data.
- the platform server 103 may also provide an algorithm model to the edge device, so that the edge device can analyze the collection data uploaded by the collection device.
- the above server may be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud service, cloud database, cloud computing, cloud function, cloud storage, network service, cloud communication, Cloud servers for basic cloud computing services such as middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms.
- the acquisition device may be a camera deployed at each target location in the industrial environment.
- the configuration parameters of the camera may change, resulting in changes in the content of the images captured by the camera.
- the focal length of the camera may change, causing changes in what the camera captures. Increasing the focal length will result in fewer targets being photographed, but each target will be sharper; decreasing the focal length will result in more targets being photographed, but each target will be blurrier.
- fine-tuning of the focal length can also result in an overall sharper or blurrier image.
- the camera may have an individual best-fit focal length.
- the shooting color and resolution of the camera may change. A change in resolution results in a change in the pixel area of the image. Changes in the color of the shot will cause the RGB values of the image to change.
- the algorithm model can be retrained.
- it takes a lot of time to re-acquire the training data and perform the training and deployment of the algorithm model, which reduces the utilization efficiency of the algorithm model.
- the present application proposes a method for processing device configuration parameters.
- the method processes the configuration parameters of the device without retraining the algorithm model, and can also improve the utilization efficiency of the algorithm model under the condition of ensuring the accuracy of the content analysis of the collected data.
- FIG. 3 schematically shows a flowchart of a method for processing a device configuration parameter according to an embodiment of the present application.
- the method may be implemented by a device configuration parameter processing apparatus.
- the device configuration parameter processing apparatus may be provided in the edge device and/or the platform server.
- the method includes:
- step S201 acquisition data of the acquisition device located at the target position is acquired.
- the collected data is configured to be input into an algorithm model for the target location, such as a content analysis model, to obtain a calculation result, such as a content analysis result of the collected data.
- an algorithm model for the target location such as a content analysis model
- step S202 based on the collected data, it is determined whether the configuration parameters of the collection device have changed.
- step S203 in response to a change in the configuration parameters of the collection device, preset configuration parameters for the target location are sent to the collection device, so as to adjust the configuration parameters of the collection device based on the preset configuration parameters .
- step S203 there is a corresponding relationship between the preset configuration parameters and the content analysis model.
- the corresponding relationship may be reflected in that the content analysis result obtained after the collection data collected by the collection device having the preset configuration parameters is input into the content analysis model has the highest content analysis accuracy. For example, if for the collection data collected by the collection device configured with the preset configuration parameters, the content analysis result output by the content analysis model has the first content analysis accuracy, and for the collection data configured with the current configuration parameters If the collected data collected by the device and the content analysis result output by the content analysis model have a second content analysis accuracy rate, the first content analysis accuracy rate is higher than the second content analysis accuracy rate.
- the above target position refers to the process position of the collecting device.
- the target location can be determined by a positioning device installed on the acquisition device itself.
- a process location refers to a location corresponding to a process step according to industrial manufacturing requirements.
- a production line includes multiple process steps, and each process step may correspond to a process position.
- the process location can be a point, or it can be a workshop location in the factory corresponding to the same process step.
- the content analysis model refers to a model constructed by a deep learning algorithm or a neural network algorithm and used to detect and identify industrial data.
- Content analysis models include, but are not limited to, defect detection models, defect rate analysis models, and installation location detection models.
- the collection data reported by the collection device at the target location can be obtained, and then it is determined whether the configuration parameters of the collection device have changed according to the collected data.
- the configuration parameters of the collection device after receiving the collection data uploaded by the collection device, it may be determined whether the configuration parameters of the collection device have changed based on the data index of the collected data or the current device identifier of the collection device that collected the collected data.
- the above data indicators are indicators that can evaluate whether the configuration parameters of the collection device are changed.
- the data indicators of the collected data may be the size, position, shape, and color of objects in the image data, or may be the resolution, brightness, contrast, sharpness, color parameters (for example, , white balance) any one or more.
- the acquisition device is generally fixed, and the surrounding environment and (eg, lighting situation) of the target location for which it is aimed are also fixed. Therefore, in the case where the configuration parameters of the collection device do not change, the data indicators in the above example generally do not change. In other words, if the above data indicators have changed, it means that the configuration parameters of the collection device have changed.
- the color parameters of the image data captured by the image acquisition device at different times for the same workstation should also be consistent.
- the color parameters of the image can be obtained through some image processing tools.
- the color parameter of the image data may be the first value. In this case, if it is found that the color parameter of the image data collected for the same station at an earlier time is a second value different from the first value, and the lighting situation of the station has not changed, it may be Indicates that the image capture device was reset due to a malfunction or other reason, so that its configuration parameters were reset.
- the position of the image acquisition device is fixed, and its position relative to the production line is also unchanged. Therefore, from the perspective of physical hardware, the position of the physical optical element in the image acquisition device relative to the workstation is also fixed. In this case, the size of the product on the production line in the image should also be the same in each frame of the image it takes.
- the configuration parameters of the optical elements inside the image capturing device may change, for example, the focal length of the image capturing device may change. This can cause the size of the product on the production line to change in the image. Therefore, it is possible to compare whether the size of the target object in the collected image data and the size of the target object in the image data collected for the same workstation at an earlier time have changed to determine whether the configuration parameters of the image capture device have changed. .
- the replacement of the collection device can also be determined by comparing the device identifiers of the collection devices.
- the acquisition data can be set to include the device ID of the acquisition device. In this way, it can be determined whether the collecting device has been replaced by comparing the current device identification of the current collecting device with the stored historical device identification of the historical collecting device for the workstation. When the current device identification is inconsistent with the historical device identification, it is determined that the configuration parameters of the collection device have changed.
- the data indicators of the collection device may be adjustable parameters of the AOI device.
- Adjustable parameters of AOI equipment include focal length, brightness, chromatic aberration, white balance, picture compression rate (or resolution), etc.
- determining whether the configuration parameters of the collection device have changed may include: comparing the collected data with the same data indicators of the historically collected data to obtain a first comparison result; and, based on the first comparison result, determining Check whether the configuration parameters of the acquisition device have changed.
- the historically collected data is data collected at historical moments for the target position. The historical moment is earlier than the data collection moment in time. In these embodiments, the historically collected data is the same data index of the temporally prior data for the same target location.
- the same data indicators refer to the same type of data indicators in the collected data and the historically collected data. For example, the white balance index of the collected data and the white balance index of the historically collected data are an example of the same data index.
- determining whether the configuration parameters of the collection device have changed includes: comparing the collected data with the same data indicators of the training data of the content analysis model to obtain a second a comparison result; and, according to the second comparison result, determining whether a configuration parameter of the collection device has changed.
- the training data for the content analysis model for the workstation is selected as the temporally prior data for the same target location.
- the data indicators of the collected data can be obtained by simply processing the collected data.
- the data metric may be the clarity of the collected data. The clarity is solely related to the position of the optical lens inside the camera of the image acquisition device. By moving the position of the optical lens, the adjustment of the sharpness can be achieved. When the optical lens moves in the direction perpendicular to the imaging plane, the distance between the lens and the imaging plane will change accordingly, resulting in changes in the sharpness of the captured image. By adjusting the position of the optical lens, the collected image can be made clearer.
- the data indicator may be a color parameter.
- the configuration parameter may be a combination of multiple parameters, such as a combination of resolution parameters, color parameters, optical zoom parameters, and the like.
- the collected data obtained at the current collection moment may be compared with the same data index as the collected data at a certain moment before the current collection moment.
- the collection data obtained at the current collection moment may be compared with the same data indicators of the training data used to train the content analysis model.
- the embodiment of the present application adjusts the configuration parameters of the acquisition device by determining the change of the configuration parameters to trigger. Compared with the manual identification of configuration parameter changes in the related art, the present application can effectively save the adjustment time of the equipment parameters, improve the equipment adjustment efficiency, and further improve the utilization efficiency of the content analysis model.
- the preset configuration parameters corresponding to the target position can be obtained, and then the preset configuration parameters are sent to the collection device, so as to adjust the collection device based on the preset configuration parameters configuration parameters.
- the term "preset configuration parameter" can be understood as follows: the content analysis result obtained based on the collection data collected by the collection device with the preset configuration parameter has a higher content analysis accuracy than the content analysis result based on the collection device with the current configuration parameter The content analysis accuracy rate of the content analysis results obtained from the collected collection data.
- the content analysis model processes and analyzes the collection data collected by collection devices with different configuration parameters, the content analysis results obtained may be different.
- the content analysis accuracy rate obtained may be higher than that obtained by inputting the collected data obtained by the collection device with other configuration parameters into the content analysis model.
- the resulting content analysis accuracy may be preset configuration parameters for such content analysis models.
- the content analysis accuracy rate is determined according to the difference between the actual situation reflected by the collected data and the analysis result output by the content analysis model after the collected data is input into the content analysis model. The smaller the difference, the more accurate the content analysis.
- the historical image data can be used as a sample, and the historical image data is the target content result as a label, which is input into the content analysis model to be constructed for training.
- target content result refers to the objective content reflected in the collected data.
- the content analysis model is a model for analyzing whether a product is defective
- the target content result refers to an objective conclusion of whether the product within the image data is defective.
- the content analysis model will output the content analysis result.
- content analysis result refers to the conclusion drawn by the content analysis model for judging whether the product in the image data is defective.
- the content analysis accuracy rate is an index for evaluating the content analysis results output by the content analysis model.
- the corresponding evaluation index is the accuracy of the defect detection result.
- the accuracy of defect detection results can reflect the matching degree between the configuration parameters of the acquisition device and the content analysis model.
- a configuration parameter classification model can be constructed first, and then the collected data is input into the configuration parameter classification model to obtain the current configuration parameter category of the collection device corresponding to the collected data; then the current configuration parameter category is compared with the preset configuration parameter category. If the current configuration parameter category of the collection device is not the preset configuration parameter category, determine that the current configuration parameter of the collection device has changed, and send the preset configuration parameter corresponding to the preset configuration parameter category to the collection device.
- current location information of the collection device may be acquired, and then preset configuration parameters corresponding to the current location information may be acquired according to the current location information.
- the preset configuration parameters are sent to the acquisition device, and the acquisition device is controlled to adjust the configuration parameters, so that the acquisition device performs data acquisition according to the adjusted configuration parameters. Then, when the collected data is input into the content analysis model for operation, accurate operation results can be obtained.
- the problem of re/retraining the content analysis model can be effectively avoided, and the operation efficiency is improved.
- sending the preset configuration parameters to the collection device may further include: converting the preset configuration parameters into control instructions corresponding to the collection equipment, and then sending the control instructions to the collection equipment, so that the collection equipment responds to the control instructions to perform parameter adjustment corresponding to the preset configuration parameters.
- the time for the acquisition device to perform parameter adjustment can be effectively saved, and the efficiency of device control can be improved.
- FIG. 4 shows an exemplary flowchart of a method for determining a preset configuration parameter provided by an embodiment of the present application. As shown in FIG. 4 , the method may be implemented by a device configuration parameter processing apparatus. The device configuration parameter processing device may be set in the edge device or the platform server. As shown in Figure 4, the method includes the following steps.
- Step S301 Obtain historical collection data for the target location, a configuration parameter category of a collection device that collects the historical collection data, and a target content result of the historical collection data, where the target content result is the source of the corresponding collection data. reflected objective content.
- Step S302 Input the second historically collected data corresponding to each configuration parameter category into the content analysis model to obtain a content analysis result of the second historically collected data corresponding to each configuration parameter category.
- Step S303 based on the content analysis result and target content result of the second historically collected data corresponding to each configuration parameter category, determine the content analysis accuracy rate of the content analysis model for the historically collected data of each configuration parameter category.
- Step S304 determining the configuration parameter category corresponding to the highest content analysis accuracy rate as a preset configuration parameter category.
- the historical collection data refers to the acquisition of multiple collection data under the conditions of different configuration parameters for the same collection device on the same workstation in the model training stage.
- a camera deployed in the same work station collects multiple image data under different configuration parameter conditions.
- Table (1) shows examples of the different configuration parameter conditions.
- the above configuration parameter classification result may be indicated by a configuration parameter category, and the configuration parameter category may be a numerical value or a letter or other symbols that can be used to identify the configuration parameter classification.
- category 1 represents the configuration parameters of the first category.
- the configuration parameters include: the resolution parameter is 720P, the zoom parameter (ie the optical zoom parameter) is 1-10x, the aperture parameter is F0.95, 3D The attribute switch parameter is on, and the color attribute is black and white.
- the historical collection data collected by the collection device configured according to the configuration parameters corresponding to the above classification 1 and the historical collection data collected by the collection device configured with the configuration parameters corresponding to the classification 2 are input into the content analysis model respectively, and different content analysis accuracy rates may be obtained. .
- the defect detection model is obtained by training based on the historical collection data collected by a certain station as a sample and the real result of whether there is a defect reflected by the historical collection data as a label.
- the production line yield is generally constant. Therefore, even if different acquisition data is input to the content analysis model, the overall product yield reflected by the content analysis results obtained should be close to the true yield.
- the content analysis accuracy rates obtained are different. This reflects that the content analysis accuracy of the content analysis model is related to the configuration parameters of the collection device. Therefore, the content analysis accuracy rate may be calculated separately for each configuration parameter category, and then the configuration parameter corresponding to the configuration parameter category with the highest content analysis accuracy rate is determined as the preset configuration parameter.
- a configuration parameter classification model may be constructed first. For example, the historical collection data for the target location and the configuration parameter category of the collection device that collects the first historical collection data may be acquired. Then, model training is performed using the historically collected data as a sample and the configuration parameter category of the collection device that collected the historically collected data as a label, so as to obtain the configuration parameter classification model. In the application stage, the collected data is input into the configuration parameter classification model, and the current configuration parameter category of the collection device can be obtained. Then, it is determined whether the current configuration parameter category of the collection device is a preset configuration parameter category. If the current configuration parameter category of the collection device is not the preset configuration parameter category, it is determined that the current configuration parameter of the collection device has changed.
- a pre-built parameter optimization model may also be invoked to optimize the preset configuration parameters to obtain optimal preset configuration parameters.
- the gradient descent optimization algorithm may be used to optimize the multiple sets of preset configuration parameters to obtain optimal adjustment ranges between the multiple sets of preset configuration parameters.
- optimal preset configuration parameters may be determined according to the optimal adjustment range.
- the embodiment of the present application can match the algorithm model by adjusting the preset configuration parameters, which can effectively improve the utilization efficiency of the algorithm model.
- the preset configuration parameters may also be optimized according to the preset adjustment range.
- the configuration parameters of the historically collected data can be classified by referring to the classification results in the above table. Then, the historically collected data is input into the configuration parameter classification model for classification to obtain the configuration parameter category associated with each historically collected data, and the historically collected data is divided into multiple subsets according to the configuration parameter category. Then for each subset, defect detection is performed using a pre-built defect detection model. Then, inspection results, such as defect rates, corresponding to each subset of historically collected data can be obtained. Then, the defect rate of each subset is compared with the real defect rate reflected by the historical collection data to obtain the content analysis accuracy of each subset, such as the defect detection accuracy.
- a configuration parameter category corresponding to the subset with the highest content analysis accuracy rate is determined as a preset configuration parameter category.
- the configuration parameter corresponding to the preset configuration parameter category is the preset configuration parameter.
- a pre-built parameter classification and aggregation model can also be invoked to perform aggregation processing on multiple parameter classification results to obtain an aggregated configuration parameter category, and then the aggregated configuration parameter category is used as the classification result.
- the configuration parameter classification model is mainly used to filter the main configuration parameters that affect target recognition.
- the complexity of parametric classification models is much lower than that of algorithmic models used for industrial inspection.
- the parameter classification model does not need to be retrained.
- the above-mentioned parameter classification model or parameter classification aggregation model can be constructed by supervised learning or unsupervised learning.
- the parameter classification aggregation model can perform aggregation processing according to the similarity between different classification results in the results output by the parameter classification model.
- multiple historical configuration parameter categories are input into a pre-built parameter classification and aggregation model to obtain the similarity between multiple historical configuration parameter categories.
- the similarity may be the similarity between any two configuration parameter categories among the multiple configuration parameter categories.
- one classification result of historical configuration parameters may be ⁇ A, B, C, D ⁇ .
- the similarity value between the classification results A and B is greater than or equal to the preset threshold, it means that the difference between the image recognition result of the configuration parameter corresponding to the classification result A and the image recognition result of the configuration parameter corresponding to the classification result B is small.
- the classification result A and the classification result B can be merged.
- classification result A or classification result B can be deleted.
- the above-mentioned preset threshold may be a preset empirical value.
- the parameter aggregation classification result is the result of aggregating classification 1 and classification 3 among the parameter classification results shown in Table (1).
- the similarity between classification 1 and classification 3 is high, and the variable magnification parameter has little influence on the detection results of the algorithm model, so classification 1 and classification 3 can be aggregated to improve the Classification processing speed, thereby improving the processing efficiency of the algorithm model.
- FIG. 5 shows an exemplary flowchart of the data analysis method provided by the embodiment of the present application.
- the method may be implemented by a data analysis device.
- the data analysis device can be set in the edge device or the platform server. As shown in Figure 5, the method includes the following steps.
- Step S401 acquiring the collection data of the collection device located at the target position
- Step S402 inputting the collection data into a configuration parameter classification model to obtain a configuration parameter category of the collection device;
- Step S403 Input the collected data into a content analysis model corresponding to the configuration parameter category of the collection device, so as to obtain a content analysis result of the collected data.
- the problem that the algorithm model training time is too long due to the change of the device configuration parameters can be solved by constructing multiple content analysis models in advance.
- the process of building a plurality of content analysis models may include:
- model training is performed using the second historically collected data as a sample and the target content result of the second historically collected data as a label to obtain the content analysis model corresponding to each configuration parameter category .
- the content analysis model and configuration parameter classification model may be stored on the platform server or edge device.
- the edge device receives the reported current collection data
- the edge device identifies whether the configuration parameters of the current collection device have changed.
- the parameter classification result of the currently collected data is obtained. Then, find and call the content analysis model corresponding to the configuration parameter category to perform operation processing on the currently collected data to obtain the content analysis result.
- the embodiment of the present application pre-builds multiple content analysis models for different configuration parameter categories, and then selects the corresponding content analysis model for processing based on the configuration parameter category corresponding to the collected data, which can solve the problem that the accuracy of the calculation result of the content analysis model is reduced. question.
- a configuration parameter classification model may also be pre-built.
- the historical collection data for the target location and the configuration parameter category of the collection device that collects the historical collection data are acquired, and then the historical collection data is used as a sample to collect the historical collection data
- the configuration parameter category of the collection device is the label, and model training is performed to obtain the configuration parameter classification model.
- a pre-built configuration parameter aggregation model may also be invoked to perform aggregation processing on multiple configuration parameter categories to obtain multiple aggregated and configuration parameter categories.
- the aggregated configuration parameter category is then used as the aforementioned configuration parameter category.
- a content analysis model corresponding to the configuration parameter category can be searched according to the configuration parameter category, so as to perform content analysis processing on the collected data.
- FIGS. 6-10 These embodiments take the acquisition device as the camera and the content analysis model as the defect detection model as an example.
- the edge gateway or edge server may adopt the open source architecture shown in FIG. 2 .
- a content analysis model component is constructed at the support service layer of the edge device, and the content analysis model component includes but is not limited to a defect detection model.
- image data under different configuration parameters are collected by the camera.
- the image data is historical acquisition data.
- the device service component that uploads these image data to the edge device.
- the configuration parameter categories such as category 1, category 2, category 3, category 4, etc. are marked for each historically collected data. Then, based on the historically collected data and its corresponding configuration parameter categories, a configuration parameter classification model can be constructed.
- the data format processing of the received collected data can be performed through the device service component.
- the data format of the collected data reported by the collection device is different from the target processing format
- the data format of the collected data is converted according to the target processing format to obtain format-converted data.
- the above-mentioned target processing format refers to the format requirements for the data by the server or gateway device that receives and processes the reported collected data.
- the XML format can be converted into JSON format.
- the core service component is used to store the acquired data after format conversion. Storage methods include but are not limited to file transfer servers, distributed file systems, etc.
- the rule engine service component obtains the format-converted image data, and then provides the format-converted data to the defect detection service component.
- the defect detection service component invokes the pre-built defect detection algorithm model to perform defect detection on the converted image data, and obtains the defect detection result and the conclusion of whether the detection result is accurate.
- the accuracy rates of the detection results corresponding to the classification results of multiple different configuration parameters are different.
- the defect detection service component reports the detection result and the accuracy of the detection result to the core service component for storage.
- the core service component establishes the correspondence between the device configuration parameters with the highest detection result accuracy and the defect detection model, and sets the device configuration parameters with the highest detection result accuracy as the preset configuration parameters.
- the defect detection service component may also directly report the detection result and the accuracy of the detection result to the device remote control service component.
- the collection device reports the collected data to the device service component.
- the device service component can perform data format conversion processing on the collected data, which is the same as the processing in the model training phase.
- the collected data after format conversion is reported to the core service component.
- the rule engine service component acquires the format-converted collection data, and then determines whether the configuration parameters of the collection device have changed based on the collected data.
- a specific way of judging may be to compare the data indicators of the collected data affecting the defect detection result.
- the device remote control service component is notified.
- the equipment remote control service component responds to the notification and obtains preset configuration parameters from the core service component, and the collection data collected by the collection equipment with the preset configuration parameters can obtain higher defect detection accuracy in the defect detection algorithm model.
- the preset configuration parameters are converted into control instructions, and the control instructions are directly sent to the acquisition device to adjust the configuration parameters of the acquisition device.
- a classification service component is also preconfigured in the edge gateway or edge server, and the classification service component is used to invoke the configuration parameter classification model, or the configuration parameter classification aggregation model.
- the configuration parameter classification model and the configuration parameter classification and aggregation model perform classification processing or classification and aggregation processing on the collected data identified and processed by the rule engine service component.
- the previous processing flow of the rule engine service component is the same as that in Figure 6.
- the rule engine service component obtains the format-converted image data, and then provides the format-converted data to the classification service component.
- the classification service component can perform classification and aggregation processing on the image data that has been manually classified in advance, so as to improve the processing speed of classification without affecting the work efficiency of the defect detection service component.
- a configuration parameter classification model can be trained and built in the following ways. Obtain historical collection data for the target location and a configuration parameter category of a collection device that collects the historical collection data; and use the historical collection data as a sample to collect the historical collection data. Model training is performed for the labels to obtain the configuration parameter classification model.
- the defect detection service component After being processed by the classification service component, the defect detection service component invokes the pre-built defect detection algorithm model to defect the converted image data, and obtains the defect detection result and the accuracy of the detection result.
- the accuracy rates of the detection results corresponding to the classification results of multiple different configuration parameters are different.
- the defect detection service component reports the detection result and the accuracy of the detection result to the core service component for storage.
- the core service component establishes the correspondence between the device configuration parameters with the highest detection result accuracy and the defect detection model, and sets the device configuration parameters with the highest detection result accuracy as the preset configuration parameters.
- the defect detection service component may also directly report the detection result and the accuracy of the detection result to the device remote control service component.
- the collection device reports the collected data to the device service component.
- the device service component can perform data format conversion processing on the collected data, which is the same as the processing in the model training phase.
- the collected data after format conversion is reported to the core service component.
- the rule engine service component acquires the format-converted collection data, and based on the collection data, determines whether the configuration parameters of the collection device have changed.
- a specific comparison method may be to compare the data indicators whose collected data affects the defect detection result.
- the device remote control service component is notified.
- the device remote control service component responds to the change of the notification, and obtains preset configuration parameters from the core service component, and the preset configuration parameters have a corresponding relationship with the defect detection algorithm model.
- the device remote control service component acquires the preset configuration parameters
- the preset configuration parameters are converted into control instructions, and the control instructions are directly sent to the acquisition device.
- the configuration parameter classification model can be constructed according to the processing method of Fig. 6 or 7.
- the established configuration parameter classification model can be delivered to edge devices.
- the collected data may be processed by the edge device.
- the collected data may be classified and processed by the platform server.
- the platform server calls the pre-built defect detection algorithm model to detect the collected data after format conversion, obtains the detection results and detection accuracy of the training data for each configuration parameter category, and establishes the configuration parameters corresponding to the detection results with the highest accuracy. Correspondence between categories and defect detection algorithm models.
- the rule engine service component of the edge device can identify the data indicators of the data reported by the collection device.
- the comparison module of the platform server compares the data indicators of the collected data with the same data indicators of the historically collected data. When the difference between the data index of the collected data and the same data index of the historically collected data exceeds a preset threshold, the platform server determines that the reported image data has changed. When it is determined that a change occurs, the platform server sends a control instruction to the edge device, and the edge device sends the preset configuration parameters to the acquisition device in the form of a control instruction.
- the above operations may be repeated multiple times, combined with the adjustment of preset configuration parameters, so that the change in the content of the image data is smaller than the preset threshold.
- the collected data is sent to the platform server.
- the platform server invokes the pre-established change identification model to identify whether the collected data has changed. When it is determined that there is a change in the collected data, the platform server generates a control instruction according to the change ratio of the collected data and preset configuration parameters. The platform server sends control commands to the edge device. The edge device sends the control command to the acquisition device.
- the method may further include:
- the edge device acquires the currently collected data collected according to the adjusted configuration parameters, and determines whether the change of the data index of the currently collected data exceeds a preset threshold. When the preset threshold is exceeded, the edge device sends the current collected data to the platform server. The platform server calls the pre-established change identification model to identify whether the current collected data has changed. When it is determined that the change of the current collected data is less than the preset threshold, the loop ends; when it is determined that the change of the current collected data is greater than or equal to the preset threshold, continue to generate control instructions according to the change ratio of the collected data and the preset configuration parameters, and repeat the preceding steps until The change of the current collected data is less than the preset threshold.
- the collection device reports the collected data to the edge device.
- the edge device identifies whether the device configuration parameters corresponding to the acquisition device have changed. When it is determined that the device configuration parameters have changed, the edge device notifies the platform server, and the platform server obtains the preset configuration parameters stored by itself, then converts the preset configuration parameters into control instructions, and sends the control instructions to the acquisition device.
- the difference between FIG. 9 and FIG. 8 is that the platform server also trains a configuration parameter classification model, and can perform aggregation processing on the classification results after the classification processing, so as to improve the speed of the classification processing.
- the platform server calls the pre-built defect detection algorithm model to detect the collected data after format conversion, and obtains the detection result and the corresponding accuracy rate of the detection result.
- the corresponding relationship between the configuration parameters corresponding to the detection result with the highest accuracy and the defect detection algorithm model is established.
- the collection device reports the collected data to the edge device.
- the edge device identifies whether the device configuration parameters corresponding to the acquisition device have changed. When it is determined that the device configuration parameters have changed, the edge device first calls the configuration parameter classification model obtained from the platform server to classify the collected data to obtain the configuration parameter category, and then the edge device notifies the platform server that the device configuration parameters have changed.
- the platform server obtains the preset configuration parameters stored by itself, then converts the preset configuration parameters into control instructions, and sends the control instructions to the acquisition device.
- FIG. 10 depicts the system flow of the data analysis method according to the present application.
- the format-converted acquisition data is obtained from the edge device through the platform server.
- the configuration parameter classification model can be set in the Algorithm Model component.
- Algorithm model components can be deployed in edge devices or in platform servers.
- the configuration parameter category may be a classification result output after being processed by the configuration parameter classification model or the configuration parameter classification aggregation model, or may be a classification result obtained through manual classification processing.
- the defect detection model can also be set in the algorithm model component. Algorithm model components can be deployed in edge devices or in platform servers.
- the platform server sends the created defect detection algorithm model corresponding to the configuration parameter category to the edge device.
- the edge device After receiving the collected data, the edge device obtains the configuration parameter category corresponding to the collected data, and then searches for the defect detection algorithm model corresponding to the configuration parameter category for detection according to the configuration parameter category. This can effectively improve the accuracy of detection results.
- a plurality of different algorithm models can be created by utilizing the computing processing capability of the platform server. It is also possible to distribute part of the algorithm model to edge devices according to the actual situation of algorithm model construction.
- the configuration parameter classification model can be arranged on edge devices.
- the parameter classification aggregation model and the defect detection algorithm model can be arranged on the platform server.
- each block in the flowchart or block diagrams may represent a module, segment, or portion of code that contains one or more logic for implementing the specified logic Executable instructions for the function.
- the functions noted in the blocks may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved.
- each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations can be implemented in dedicated hardware-based systems that perform the specified functions or operations , or can be implemented in a combination of dedicated hardware and computer instructions.
- FIG. 11 shows a schematic structural diagram of a device configuration parameter processing apparatus provided by an embodiment of the present application.
- the device configuration parameter processing apparatus can be configured on an edge device or a platform server, that is, through a memory, a processor in the edge device or the platform server, and a computer program stored in the memory and run on the processor, when the program is executed, the Implement the method described in Figure 3.
- the device includes:
- Acquisition data acquisition module 1001 which is configured to acquire the acquisition data of the acquisition device located at the target location, wherein the acquired data is configured to be input into a content analysis model for the target location to obtain a content analysis result of the acquired data ;
- a configuration parameter change determination module 1002 which is configured to determine whether the current configuration parameter of the collection device has changed based on the collected data
- a preset configuration parameter sending module 1003 which is configured to send preset configuration parameters for the target location to the collection device in response to changes in the current configuration parameters of the collection device, so as to be based on the preset configuration parameters adjusting the current configuration parameter of the collection device; wherein, for the collection data collected by the collection device configured with the preset configuration parameter, the content analysis result output by the content analysis model has a first content analysis accuracy, For the collection data collected by the collection device configured with the current configuration parameters, the content analysis result output by the content analysis model has a second content analysis accuracy, and the first content analysis accuracy is higher than the second content Analysis accuracy.
- the configuration parameter change determination module 1002 is further configured to compare the same data indicators of the collected data and historically collected data to obtain a first comparison result, wherein the historically collected data is for the target location at Data collected at a historical moment, which is earlier than the data collection moment in time; and, based on the first comparison result, determining whether the current configuration parameter of the collection device has changed.
- the configuration parameter change determination module 1002 is further configured to compare the same data metrics of the collected data and the training data of the content analysis model to obtain a second comparison result, wherein the training data is used for data for training the content analysis model; and, based on the second comparison result, determining whether a current configuration parameter of the collection device has changed.
- the apparatus further includes a preset configuration parameter determination module.
- the preset configuration parameter determination module is configured to: acquire historical collection data for the target location, a configuration parameter category of a collection device that collects the historical collection data, and a target content result of the historical collection data, the target content result is the objective content reflected by the corresponding collection data; input the historical collection data corresponding to each configuration parameter category into the content analysis model to obtain the content analysis result of the historical collection data corresponding to each configuration parameter category ; Based on the content analysis results and target content results of the historically collected data corresponding to each configuration parameter category, determine the content analysis accuracy rate of the content analysis model for the historically collected data of each configuration parameter category; The configuration parameter category corresponding to the accuracy rate is determined as the preset configuration parameter category.
- the configuration parameter change determination module 1002 is further configured to input the collected data into a configuration parameter classification model to obtain the current configuration parameter category of the collection device; determine the current configuration parameter category of the collection device Whether it is a preset configuration parameter category; in response to the current configuration parameter category of the collection device not being the preset configuration parameter category, it is determined that the current configuration parameter of the collection device has changed.
- the preset configuration parameter sending module 1003 is configured to send the preset configuration parameters corresponding to the preset configuration parameter categories to the collection device.
- the apparatus further includes a configuration parameter classification model building module configured to obtain historically collected data for the target location and a configuration parameter category of a collection device that collected the historically collected data; The historically collected data is used as a sample, and model training is performed with the configuration parameter category of the collection device that collected the historically collected data as a label, so as to obtain the configuration parameter classification model.
- a configuration parameter classification model building module configured to obtain historically collected data for the target location and a configuration parameter category of a collection device that collected the historically collected data; The historically collected data is used as a sample, and model training is performed with the configuration parameter category of the collection device that collected the historically collected data as a label, so as to obtain the configuration parameter classification model.
- the configuration parameter classification model building module may also be configured to, after acquiring the configuration parameter category of the collection device that collects the historical collection data, input the configuration parameter category of the collection device that collects the historical collection data into the configuration parameter category aggregation model, to obtain an aggregated configuration parameter category; and, using the historically collected data as a sample and the aggregated configuration parameter category as a label, perform model training to obtain the configuration parameter classification model.
- the apparatus further includes a content analysis model building module configured to obtain historical collection data for the target location and target content results of the historical collection data, the target content results being corresponding collection data The objective content reflected; take the historically collected data as a sample and use the target content result of the historically collected data as a label to perform model training to obtain the content analysis model.
- a content analysis model building module configured to obtain historical collection data for the target location and target content results of the historical collection data, the target content results being corresponding collection data The objective content reflected; take the historically collected data as a sample and use the target content result of the historically collected data as a label to perform model training to obtain the content analysis model.
- FIG. 12 shows a schematic structural diagram of a data analysis apparatus provided by an embodiment of the present application.
- the data analysis apparatus can be configured on an edge device or a platform server, that is, through a memory, a processor in the edge device or the platform server, and a computer program stored in the memory and run on the processor, when the program is executed, the graph is realized. 5 Describe the method.
- the device includes:
- a configuration parameter category determination module 1102 which is configured to input the collected data into a configuration parameter classification model to obtain a configuration parameter category of the collection device;
- the content analysis module 1103 is configured to input the collected data into a content analysis model corresponding to the configuration parameter category of the collection device, so as to obtain a content analysis result of the collected data.
- the problem of low utilization of a single algorithm model caused by multiple training of a single algorithm model is solved by pre-configuring multiple algorithm models.
- FIG. 13 shows a schematic structural diagram of a computing device provided by an embodiment of the present application.
- the system includes: an edge device 1201 and a platform server 1202 as described in Figure 11 or 12, a communication connection is established between the edge device and the platform server, and the platform server 1202 is configured to provide the edge device with pre-built algorithm models and parameter classification models.
- FIG. 14 shows a schematic structural diagram of a computer system of a device or a server provided by an embodiment of the present application.
- the computer system includes a central processing unit (CPU) 1301, which can be processed according to a program stored in a read only memory (ROM) 1302 or a program loaded from a storage section 908 into a random access memory (RAM) 1303 Various appropriate actions and processes are performed.
- ROM read only memory
- RAM random access memory
- various programs and data required for system operation are also stored.
- the CPU 1301, the ROM 1302, and the RAM 1303 are connected to each other through a bus 1304.
- An input/output (I/O) interface 1305 is also connected to bus 1304 .
- the following components are connected to the I/O interface 1305: an input section 1306 including a keyboard, a mouse, etc.; an output section 1307 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1308 including a hard disk, etc. ; and a communication section 1309 including a network interface card such as a LAN card, a modem, and the like.
- the communication section 1309 performs communication processing via a network such as the Internet.
- Drivers 1310 are also connected to I/O interface 1305 as needed.
- a removable medium 1311 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is mounted on the drive 1310 as needed so that a computer program read therefrom is installed into the storage section 1308 as needed.
- the process described above with reference to the flowcharts of FIG. 3 or 5 may be implemented as a computer software program.
- embodiments of the present disclosure include a computer program product comprising a computer program carried on a machine-readable medium, the computer program containing program code for performing the method illustrated in the flowchart.
- the computer program may be downloaded and installed from the network via the communication portion 1309, and/or installed from the removable medium 1311.
- CPU central processing unit
- the computer-readable medium shown in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the above two.
- the computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or a combination of any of the above. More specific examples of computer readable storage media may include, but are not limited to, electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable Programmable read only memory (EPROM or flash memory), fiber optics, portable compact disk read only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.
- a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.
- a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code therein. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing.
- a computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device .
- Program code embodied on a computer readable medium may be transmitted using any suitable medium including, but not limited to, wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
- the application also provides a computer program product or computer program.
- the computer program product or computer program includes computer instructions.
- the computer instructions are stored in a computer-readable storage medium.
- a processor of the computing device reads the computer instructions from a computer-readable storage medium.
- the processor executes the computer instructions, so that the computing device executes the device configuration parameter processing method and the data analysis method provided in the above-mentioned various optional implementation manners.
- each block in the flowchart or block diagrams may represent a module, segment, or portion of code that contains one or more functions for implementing the specified logical function(s) executable instructions.
- the functions noted in the blocks may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved.
- each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations can be implemented in dedicated hardware-based systems that perform the specified functions or operations , or can be implemented in a combination of dedicated hardware and computer instructions.
- the units or modules involved in the embodiments of the present application may be implemented in a software manner, and may also be implemented in a hardware manner.
- the described unit or module can also be set in the processor, for example, it can be described as: a processor includes a first collected data obtaining unit, a determining unit, a parameter obtaining unit and a sending unit.
- a processor includes a first collected data obtaining unit, a determining unit, a parameter obtaining unit and a sending unit.
- the names of these units or modules do not constitute a limitation on the unit or module itself under certain circumstances.
- the service subscription module can also be described as "a unit used to obtain the collection data reported by the collection device at the target location. ".
- the present application also provides a computer-readable storage medium.
- the computer-readable storage medium may be included in the electronic device described in the above-mentioned embodiments; in electronic equipment.
- the above-mentioned computer-readable storage medium stores one or more programs, when the above-mentioned programs are used by one or more processors to execute the device parameter processing method described in this application.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Automation & Control Theory (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
Description
Claims (20)
- 一种设备配置参数处理方法,包括:获取位于目标位置的采集设备的采集数据,其中所述采集数据配置成输入到针对所述目标位置的内容分析模型中,以得到所述采集数据的内容分析结果;基于所述采集数据,确定所述采集设备的当前配置参数是否发生变化;响应于所述采集设备的当前配置参数发生变化,将针对所述目标位置的预设配置参数发送至所述采集设备,以基于所述预设配置参数调整所述采集设备的所述当前配置参数;其中,在所述内容分析模型的输入是配置有所述预设配置参数的采集设备的采集数据的情况下,所述内容分析模型输出的内容分析结果具有第一内容分析准确率,在在所述内容分析模型的输入是配置有所述当前配置参数的采集设备的采集数据的情况下,所述内容分析模型输出的内容分析结果具有第二内容分析准确率,其中所述第一内容分析准确率高于所述第二内容分析准确率。
- 根据权利要求1所述的方法,其中,基于所述采集数据,确定所述采集设备的当前配置参数是否发生变化包括:比较所述采集数据和历史采集数据的相同数据指标,以得到第一比较结果,其中所述历史采集数据是针对所述目标位置在历史时刻采集的数据,所述历史时刻在时间上早于所述数据采集时刻,所述数据指标是所述采集数据的随着采集设备的配置参数的变化而变化的指标;以及,基于所述第一比较结果,确定所述采集设备的当前配置参数是否发生变化。
- 根据权利要求1所述的方法,其中,基于所述采集数据,确定所述采集设备的当前配置参数是否发生变化包括:比较所述采集数据和所述内容分析模型的训练数据的相同数据指标,以得到第二比较结果,其中所述训练数据是用于训练所述内容分析模型的数据,所述数据指标是所述采集数据的随着采集设备的配置参数的变化而变化的指标;以及,基于所述第二比较结果,确定所述采集设备的当前配置参数是否发生变化。
- 根据权利要求1所述的方法,其中,所述采集数据包括所述采集设备的当前设备标识,并且,基于所述采集数据,确定所述采集设备的当前配置参数是否发生变化包括:从所述采集数据中,识别出所述采集设备的设备标识;以及,基于所述采集设备的设备标识与所述目标位置的历史设备标识,确定所述采集设备的当前配置参数是否发生变化。
- 根据权利要求1所述的方法,还包括:在获取位于目标位置的采集设备的采集数据之后,对所述采集数据进行数据格式转换。
- 根据权利要求1所述的方法,其中,基于所述采集数据,确定所述采集设备的当前配置参数是否发生变化包括:将所述采集数据输入到配置参数分类模型,以得到所述采集设备的当前配置参数类别;确定所述采集设备的当前配置参数类别是否为预设配置参数类别;响应于所述采集设备的当前配置参数类别不是所述预设配置参数类别,确定所述采集设备的当前配置参数发生变化;并且,将针对所述目标位置的预设配置参数发送至所述采集设备包括:将所述预设配置参数类别所对应的预设配置参数发送至所述采集设备。
- 根据权利要求6所述的方法,还包括:获取针对所述目标位置的第一历史采集数据以及采集所述第一历史采集数据的采集设备的配置参数类别;并且,以所述第一历史采集数据为样本、以采集所述第一历史采集数据的采集设备的配置参数类别为标签进行模型训练,以得到所述配置参数分类模型。
- 根据权利要求7所述的方法,还包括:在获取采集所述第一历史采集数据的采集设备的配置参数类别之后,将采集所述第一历史采集数据的采集设备的配置参数类别输入到 配置参数类别聚合模型,以得到经聚合的配置参数类别;并且,以所述第一历史采集数据为样本、以采集所述第一历史采集数据的采集设备的配置参数类别为标签进行模型训练包括:以所述第一历史采集数据为样本、以所述经聚合的配置参数类别为标签进行模型训练,以得到所述配置参数分类模型。
- 根据权利要求6所述的方法,还包括:获取针对所述目标位置的第二历史采集数据、采集所述第二历史采集数据的采集设备的配置参数类别、以及所述第二历史采集数据的目标内容结果,所述目标内容结果是所述第二历史采集数据所反映的客观内容;将每个配置参数类别对应的所述第二历史采集数据输入到所述内容分析模型,以得到每个配置参数类别对应的所述第二历史采集数据的内容分析结果;基于每个配置参数类别对应的所述第二历史采集数据的内容分析结果和目标内容结果,确定所述内容分析模型对于每个配置参数类别的第二历史采集数据的内容分析准确率;将最高的内容分析准确率对应的配置参数类别确定为所述预设配置参数类别。
- 根据权利要求6所述的方法,还包括:获取针对所述目标位置的第三历史采集数据以及所述第三历史采集数据的目标内容结果,所述目标内容结果是所述第三历史采集数据所反映的客观内容;以所述第三历史采集数据为样本、以所述第三历史采集数据的目标内容结果为标签进行模型训练,以得到所述内容分析模型。
- 根据权利要求1-10中的任一项所述的方法,其中,所述方法由边缘设备执行,所述边缘设备包括规则引擎服务组件和设备远程控制服务组件,并且,基于所述采集数据,确定所述采集设备的当前配置参数是否发生变化由所述规则引擎服务组件执行,并且,将针对所述目标位置的预设配置参数发送至所述采集设备由所述设备远程控制服务组件执行。
- 根据权利要求1-10中的任一项所述的方法,其中,所述方法由边缘设备和平台服务器执行,并且,基于所述采集数据,确定所述采集设备的当前配置参数是否发生变化由所述边缘设备执行,并且,将针对所述目标位置的预设配置参数发送至所述采集设备由所述平台服务器执行。
- 一种数据分析方法,包括:获取位于目标位置的采集设备的采集数据;将所述采集数据输入到配置参数分类模型,以得到所述采集设备的配置参数类别;将所述采集数据输入到与所述采集设备的配置参数类别对应的内容分析模型,以得到所述采集数据的内容分析结果。
- 根据权利要求13所述的方法,还包括:获取针对所述目标位置的第一历史采集数据以及采集所述第一历史采集数据的采集设备的配置参数类别;以所述第一历史采集数据为样本、以采集所述第一历史采集数据的采集设备的配置参数类别为标签进行模型训练,以得到所述配置参数分类模型。
- 根据权利要求13所述的方法,还包括:获取针对所述目标位置的第二历史采集数据、采集所述第二历史采集数据的采集设备的配置参数类别、以及所述第二历史采集数据的目标内容结果,所述目标内容结果是所述第二历史采集数据所反映的客观内容;针对每个配置参数类别,以所述第二历史采集数据为样本、以所述第二历史采集数据的目标内容结果为标签进行模型训练,以得到每个配置参数类别对应的所述内容分析模型。
- 根据权利要求13所述的方法,其中,将所述采集数据输入到配置参数分类模型,以得到所述采集设备的配置参数类别以及将所述采集数据输入到与所述采集设备的配置参数类别对应的内容分析模型,以得到所述采集数据的内容分析结果由算法模型组件执行,其中所述算法模型组件布置在边缘设备或平台服务器中。
- 一种设备配置参数处理装置,包括:采集数据获取模块,其配置成获取位于目标位置的采集设备的采集数据,其中所述采集数据配置成输入到针对所述目标位置的内容分析模型中,以得到所述采集数据的内容分析结果;配置参数变化确定模块,其配置成基于所述采集数据,确定所述采集设备的当前配置参数是否发生变化;预设配置参数发送模块,其配置成响应于所述采集设备的当前配置参数发生变化,将针对所述目标位置的预设配置参数发送至所述采集设备,以基于所述预设配置参数调整所述采集设备的所述当前配置参数;其中,在所述内容分析模型的输入是配置有所述预设配置参数的采集设备的采集数据的情况下,所述内容分析模型输出的内容分析结果具有第一内容分析准确率,在在所述内容分析模型的输入是配置有所述当前配置参数的采集设备的采集数据的情况下,所述内容分析模型输出的内容分析结果具有第二内容分析准确率,其中所述第一内容分析准确率高于所述第二内容分析准确率。
- 一种计算设备,包括:存储器,其被配置成存储计算机可执行指令;处理器,其被配置成当所述计算机可执行指令被所述处理器执行时,实现如权利要求1-16中的任一项所述的方法。
- 一种计算机可读存储介质,其上存储有计算机可执行指令,当所述计算机可执行指令被执行时,实现如权利要求1-16中的任一项所述的方法。
- 一种计算机程序产品,包括计算机可执行指令,其中所述计算机可执行指令被处理器执行时执行根据权利要求1-16中任一项所述的方法。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US18/258,634 US20240048441A1 (en) | 2020-12-25 | 2021-12-23 | Method and apparatus for device configuration parameter processing, method and apparatus for data analysis, computing device, computer-readable storage medium and computer program product |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011565492.0A CN114697212A (zh) | 2020-12-25 | 2020-12-25 | 设备参数处理方法、设备、系统及介质 |
CN202011565492.0 | 2020-12-25 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022135539A1 true WO2022135539A1 (zh) | 2022-06-30 |
Family
ID=82129573
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2021/140929 WO2022135539A1 (zh) | 2020-12-25 | 2021-12-23 | 设备配置参数处理方法和装置、数据分析方法和装置、计算设备、计算机可读存储介质、以及计算机程序产品 |
Country Status (3)
Country | Link |
---|---|
US (1) | US20240048441A1 (zh) |
CN (1) | CN114697212A (zh) |
WO (1) | WO2022135539A1 (zh) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115456459A (zh) * | 2022-09-30 | 2022-12-09 | 浙江中泽精密科技有限公司 | 一种新能源动力电池盖板片加工工艺方法及系统 |
CN115865670A (zh) * | 2023-02-27 | 2023-03-28 | 灵长智能科技(杭州)有限公司 | 基于内核调优的web安全网关并发性能调节方法及装置 |
CN116224879A (zh) * | 2023-03-22 | 2023-06-06 | 佛山市众合科技有限公司 | 一种基于云计算的工业数据处理系统 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106355319A (zh) * | 2016-08-26 | 2017-01-25 | 南京安全无忧网络科技有限公司 | 一种安全风险评估系统和方法 |
CN107193967A (zh) * | 2017-05-25 | 2017-09-22 | 南开大学 | 一种多源异构行业领域大数据处理全链路解决方案 |
CN107431751A (zh) * | 2016-02-03 | 2017-12-01 | 华为技术有限公司 | 采集图像的方法、装置和系统 |
CN111177095A (zh) * | 2019-12-10 | 2020-05-19 | 中移(杭州)信息技术有限公司 | 日志分析方法、装置、计算机设备及存储介质 |
US20200260962A1 (en) * | 2015-11-09 | 2020-08-20 | Magniware Ltd. | System and methods for acquisition and analysis of health data |
-
2020
- 2020-12-25 CN CN202011565492.0A patent/CN114697212A/zh active Pending
-
2021
- 2021-12-23 WO PCT/CN2021/140929 patent/WO2022135539A1/zh active Application Filing
- 2021-12-23 US US18/258,634 patent/US20240048441A1/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20200260962A1 (en) * | 2015-11-09 | 2020-08-20 | Magniware Ltd. | System and methods for acquisition and analysis of health data |
CN107431751A (zh) * | 2016-02-03 | 2017-12-01 | 华为技术有限公司 | 采集图像的方法、装置和系统 |
CN106355319A (zh) * | 2016-08-26 | 2017-01-25 | 南京安全无忧网络科技有限公司 | 一种安全风险评估系统和方法 |
CN107193967A (zh) * | 2017-05-25 | 2017-09-22 | 南开大学 | 一种多源异构行业领域大数据处理全链路解决方案 |
CN111177095A (zh) * | 2019-12-10 | 2020-05-19 | 中移(杭州)信息技术有限公司 | 日志分析方法、装置、计算机设备及存储介质 |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115456459A (zh) * | 2022-09-30 | 2022-12-09 | 浙江中泽精密科技有限公司 | 一种新能源动力电池盖板片加工工艺方法及系统 |
CN115456459B (zh) * | 2022-09-30 | 2023-05-05 | 浙江中泽精密科技有限公司 | 一种新能源动力电池盖板片加工工艺方法及系统 |
CN115865670A (zh) * | 2023-02-27 | 2023-03-28 | 灵长智能科技(杭州)有限公司 | 基于内核调优的web安全网关并发性能调节方法及装置 |
CN115865670B (zh) * | 2023-02-27 | 2023-06-16 | 灵长智能科技(杭州)有限公司 | 基于内核调优的web安全网关并发性能调节方法及装置 |
CN116224879A (zh) * | 2023-03-22 | 2023-06-06 | 佛山市众合科技有限公司 | 一种基于云计算的工业数据处理系统 |
Also Published As
Publication number | Publication date |
---|---|
US20240048441A1 (en) | 2024-02-08 |
CN114697212A (zh) | 2022-07-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2022135539A1 (zh) | 设备配置参数处理方法和装置、数据分析方法和装置、计算设备、计算机可读存储介质、以及计算机程序产品 | |
US20240202893A1 (en) | Method and device for detecting defect, storage medium and electronic device | |
JP7393515B2 (ja) | 分散型製品欠陥分析システム、方法及びコンピュータ可読記憶媒体 | |
CN111814850A (zh) | 缺陷检测模型训练方法、缺陷检测方法及相关装置 | |
JP2014533349A (ja) | ラスタ画像のエッジ検出によるベルト摩耗の測定 | |
CN108932273B (zh) | 图片筛选方法及装置 | |
US11288101B2 (en) | Method and system for auto-setting of image acquisition and processing modules and of sharing resources in large scale video systems | |
CN111611889B (zh) | 基于改进卷积神经网络的农田微型虫害识别装置 | |
WO2020007119A1 (zh) | 显示屏外围电路检测方法、装置、电子设备及存储介质 | |
CN113052039B (zh) | 一种交通路网行人密度检测的方法、系统及服务器 | |
CN115861210B (zh) | 一种基于孪生网络的变电站设备异常检测方法和系统 | |
JP7429756B2 (ja) | 画像処理方法、装置、電子機器、記憶媒体及びコンピュータプログラム | |
CN108764278A (zh) | 一种基于视觉的自学习工业智能检测系统及方法 | |
CN113037783A (zh) | 一种异常行为检测方法及系统 | |
CN114998679A (zh) | 深度学习模型的在线训练方法、装置、设备及存储介质 | |
WO2019223763A1 (zh) | 图像检测方法、装置、设备及介质,图案化控制系统及方法 | |
Bharti et al. | Edge-enabled federated learning for vision based product quality inspection | |
US11068734B2 (en) | Client terminal for performing hybrid machine vision and method thereof | |
CN117351472A (zh) | 烟叶信息检测方法、装置及电子设备 | |
WO2024000372A1 (zh) | 一种缺陷检测的方法和装置 | |
WO2020124448A1 (en) | Systems and methods for video surveillance | |
CN115147752A (zh) | 一种视频分析方法、装置及计算机设备 | |
CN112949490A (zh) | 设备动作检测方法、装置、电子设备及可读存储介质 | |
CN112597788A (zh) | 目标测定方法、装置、电子设备和计算机可读介质 | |
CN111209057A (zh) | 一种快速调整感知终端的方法以及装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21909534 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 18258634 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 24.10.2023) |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 21909534 Country of ref document: EP Kind code of ref document: A1 |