46 results about "Edge analytics" patented technology

Disclosed is a method and system for processing images from an aerial imaging device. A moving vehicle analysis system receives images from an aerial imaging device. The system may perform edge analysis in the images to identify a pairs of edges corresponding to a road. The system may identify pixel blobs in the images including adjacent pixels matching each other based on a pixel attribute. The system uses a machine learning model for generating an output identifying moving vehicles in the images. The system determines a count of the moving vehicles captured by the images, where each moving vehicle is associated with corresponding pixel blobs.

A system and method for accelerating web page delivery is disclosed in one embodiment. Web content requests are made to an edge server of a first point of presence (POP) of a content delivery network (CDN). The web content has embedded resource links. The first POP can rewrite the embedded resource links to route requests for the embedded resource links to any POP in the CDN or even the origin server. In some embodiments, the first POP can decide if the first POP and/or another POP referenced in a rewritten embedded resource link should cache and/or accelerate the resource referenced in the embedded resource link.

A method, system, and program for selectively testing output signals of an integrated circuit. The system comprising a mask generation file specifying output signals and test cycles and a verification module to check simulation output data and generate test pattern data. The verification module further comprising an extractor routine which receives the mask generation file and processes the simulation output data such that the test pattern data is coded to mask the specified output signals at the specified test cycles. The verification module further comprising margin analysis for determining if margin times of the coded test patterns fall below a minimum margin time setting.

A computer implemented method, computer usable program code, and a data processing system for selecting a candidate implementation of a virtual overridden method for inlining into a calling method. A determination as to which implementation of a virtual overridden method to inline is made based on its relative “hotness” compared to the other implementations of the same method. The relative hotness can be inferred from the invocation count and sampling count that the virtual machine and the just-in-time compiler already collect for other purposes, that is, without collecting and storing of call-edge profiling information. When a method is being compiled and it is identified that the method contains a call to an overridden method, a candidate for inlining from among the implementations of the overridden method is selected based on relative hotness values. The candidate implementation of the overridden method is then inlined, with a guard, into the calling method.

A method, computer program product, and computer system are disclosed. The method includes determining probability of high data flow event, memory requirements for the high data flow and time to high data flow event by using pattern recognition based on natural language Processing (NLP), machine learning and statistical algorithms. The method applies a configuration to the edge device and network system when the probability of high data flow event is above a predetermined threshold for handling the high data flow event.

The invention discloses an industrial intelligent controller capable of deploying an AI model dynamically, relating to the technical field of industrial intelligent controller. The industrial intelligent controller comprises an EDGE BOX end and a control center, wherein the EDGE BOX end and the control center transmit or communicate with each other through a multi-path serial port, an Ethernet interface, WiFi or 5G netowrk, the EDGE BOX end is communicated connection with a production machine table through a serial port, the EDGE BOX based on an edge calculation controller of an X86 processorintegrates the multi-path serial port, the Ethernet interface, WIFI and the 5G wireless communication module, and the control center, a K8S control end the EDGE BOX end are used for collecting data ofthe production machine table, sending a control command to the control center, and deploying an AI operation model and edge calculation nodes of an operation component dynamically through the controlcenter. The industrial intelligent controller capable of deploying the AI model dynamically can perform a complex algorithm and a device model in local and can perform intelligent edge analysis and control, thereby having better effect in real time performance.

The invention provides a demand side carbon flow edge analysis method, terminal and system. The system comprises a carbon flow monitoring terminal connected with a carbon flow edge terminal, a demand side energy consumption terminal and a carbon flow master station; the carbon flow edge terminal collects field data of each energy consumption end of a demand side energy consumption unit; carbon flow accounting, carbon emission analysis and energy efficiency analysis are carried out according to energy types on the basis of all field data; a regulation and control command is formed based on the carbon flow accounting result, the carbon emission analysis result and a pre-stored regulation and control strategy; the regulation and control command is issued to an energy consumption end and/or a carbon flow monitoring terminal; and the field data, the carbon flow accounting result, the carbon emission analysis result and the energy efficiency analysis result are uploaded to a carbon flow master station. The carbon flow edge terminal can perform edge calculation and analysis on the field data of the energy consumption end, so that the local application response efficiency can be enhanced, and the energy consumption end can be subjected to optimization control.

The invention provides an electric power safety management video intelligent shunting device and method based on mobile edge calculation. The device comprises a shunting strategy optimization evaluation module, and a scene detection module, a dynamic change area segmentation module, a model calculation overhead evaluation module, a model energy consumption overhead evaluation module, an equipmentstate monitoring unit, a local analysis module and a compression shunting module which are connected with the shunting strategy optimization evaluation module; the shunting strategy optimization evaluation module is used for storing scene detection and region segmentation pictures, determining and optimizing a shunting strategy, and shunting the shunting strategy to the local analysis module, an adjacent edge analysis unit or a video processing cloud analysis unit for intelligent identification. According to the invention, the shunting strategy can be adaptively adjusted according to the operation state of the edge equipment; the convergence of multiple paths of videos, the compression of shunting contents, the optimization of the shunting strategy and the scheduling of shunting tasks arerealized through moving the edge equipment, so the video time delay is solved, the terminal energy consumption is balanced, the traffic overhead is reduced, and the safety management efficiency is improved.

The invention relates to the technical field of image data processing, in particular to a municipal environmental sanitation sweeping garbage truck sweeping method based on road garbage classification. According to the method, the occlusion area in the road garbage image is determined according to the category probability identified by the garbage category identification neural network. And carrying out super-pixel segmentation on the shielding region to obtain a plurality of super-pixel blocks. And expanding the boundary line of the superpixel block to obtain an edge analysis area. Obtaining a plurality of candidate boundary lines in the edge analysis region, obtaining confidence coefficients according to the segmentation capability and pixel value distribution uniformity of the candidate boundary lines, taking the candidate boundary line with the maximum confidence coefficient as a segmentation edge line, obtaining a single junk image, and carrying out analysis by reusing the neural network, thereby achieving the accurate classification of junk. And controlling the cleaning garbage truck to perform targeted cleaning according to the garbage category. The garbage is accurately classified, the sweeping garbage truck is controlled to conduct targeted sweeping according to the classification result, the sweeping efficiency is improved, and meanwhile power consumption is reduced.

Systems and methods described herein are directed to changing models and analytics algorithms on an edge node from a core server, which can be useful in situations such as optimized factories when the edge node is physically close to data sources such as within a factory plant. The core server or cloud runs analytics algorithms and models concurrently with data received from edge nodes, and replaces the edge node with more accurate analytics algorithms and models as applicable.

The invention relates to a digital online measurement method for improving the measurement accuracy of a hole expansion ratio. The method comprises the following steps: an image captured through a camera in real time is transmitted to a computer control system in real time; the computer analyzes and processes the edge state of an expanded hole in the image through the image processing method, andautomatically judges whether the expanded hole reaches a limit state or not through setting of a certain threshold value; and meanwhile the computer analyzes the determined limit moment correspondingto the hole edge on a picture to obtain a real-time aperture of the expanded hole. By use of the method, the automation of crack judgment is implemented, the influence of human factors is avoided, andthe result is relatively stable; meanwhile the measured aperture is based on a boundary recognition algorithm of image processing, thereby avoiding the human measurement error; and the aperture is measured online but is not obtained by taking down a sample and rebounding, thus the measurement precision is improved.

The invention discloses an intelligent lighting control system. The system comprises a meteorological data collection module used for collecting meteorological data of a controlled lighting area, an illuminance sensor used for collecting illuminance of a lighting lamp, an Internet of Things controller, an intelligent gateway and a cloud platform. And the cloud platform is in communication connection with the meteorological data collection module and the intelligent gateway, generates a control instruction according to the meteorological data and the illuminance, and sends the control instruction to the intelligent gateway, and the control instruction is transmitted to the Internet of Things controller by the intelligent gateway, thereby realizing adjustment of the illuminance and/or colortemperature of the illuminating lamp. According to the method, strategy decision is carried out by combining meteorological data collection and an edge analysis algorithm, road safety lighting under the abnormal severe weather condition is effectively improved, and basic full coverage of road lighting encountering conditions is achieved. Compared with the prior art, the cloud computing analysis algorithm is adopted for strategy decision making, and the road lighting control accuracy is effectively improved.

The invention discloses a scene-based edge analysis algorithm issuing method and device and a storage medium. The scene-based edge analysis algorithm issuing method comprises the following steps: acquiring an existing video stream or picture of an installed camera; performing scene recognition on the acquired video streams or photos by adopting a neural network model, and outputting a to-be-classified scene; and comparing the to-be-classified scene with a target scene in the platform side scene library, and configuring an edge analysis algorithm corresponding to the target scene in the scene library according to a comparison result. According to the invention, through pre-analysis of the edge analysis host and comparison of the scene library at the platform side, the most appropriate and most improper algorithm results are obtained, and through combination with simple configuration of personnel, the most reasonable scene algorithm application is obtained. The method has the greatest advantages that the complexity of personnel operation is simplified by utilizing the method, the most reasonable scheme recommendation is given, the use pain point of a customer is overcome, and the most rationalization of algorithm configuration is realized in combination with scene-based pre-analysis.

The invention relates to data processing, in particular to an embedded industrial personal computer based on an edge computing system, comprising a controller, wherein the controller is connected witha data transmission module for creating a connection transmission protocol between the controller and an industrial device, the data transmission module establishes a connection with the industrial device via a real-time bus module, the controller is connected with a data driving module for importing and driving data converted by the data transmission module into the controller according to related contents of the connection transmission protocol, the controller is connected with an edge analysis module for performing edge computing on the data converted by the data transmission module, and the controller is connected with a state sensing module for sensing the state of the industrial device according to the data converted by the data transmission module; and by adopting the technical scheme provided by the invention, the defects that a fault model cannot be predicted according to the running state of the device, and that the original running data of the device cannot be stored or queried in the prior art can be effectively overcome.

The invention discloses a machine-oriented universal coding method, which comprises the following steps of: a compression coding stage: for different machine intelligent analysis tasks, compressing corresponding video or image data by adopting an independent task mode and/or a multi-task sharing mode to obtain corresponding feature information; putting the feature information into a code stream corresponding to the video or image data, and giving annotation information required in decoding in combination with a compression mode, thereby obtaining a feature code stream; and a decoding stage: reading header information of the feature code stream, determining a compression mode used in the compression stage according to annotation information in the header information so as to read the feature information in a corresponding mode, and taking the feature information as input of a machine intelligent analysis task to obtain an analysis result. Coding of video/image feature information required by each task can be realized, so that the intelligent task analysis realization efficiency is improved, the transmission pressure is reduced, and future possible requirements of edge analysis calculation and the like are supported.

The invention relates to a screen defect classification method, electronic equipment and a storage medium. The screen defect classification method comprises the steps: preprocessing an input screen defect image to obtain a defect sample image, wherein preprocessing comprises filtering, threshold segmentation, defect area and edge analysis, scale adjustment and downsampling. The defect sample image comprises a defect ROI region, the size of the defect sample image accords with the input image size of the screen defect detection network, and the defect sample image is classified through the screen defect detection network to obtain the defect type of the screen defect image. According to the method and the device, the defect sample image containing the defect ROI can be quickly and accurately acquired by combining the preprocessing and the screen defect detection network, and the defect sample image with the accurate defect ROI can improve the classification accuracy of the screen defect detection network. In addition, the problem that parameters and operation steps need to be adjusted according to different defect types in a traditional algorithm can be solved.

The invention discloses a residential user power consumption characteristic statistics method and system based on edge calculation. The method comprises the following steps: 1) performing data acquisition; 2) collecting power consumption, storing data and analyzing data edges by adopting a unified cooperative instruction; 3) enabling the data storage module to perform data storage period configuration; 4) enabling the data storage module to respond to a far-end (cloud platform system) data demand instruction; 5) data analysis: enabling a data edge analysis unit to carrty out statistical analysis on the data in the data storage module; 6) generating a report, and after the data edge analysis unit completes statistical analysis, and generating a digital statistical report according to a prefabricable characteristic analysis report generation period. The system comprises a data acquisition unit, a data storage unit, a data edge analysis unit, a cloud platform system, a wireless transmission module and a resident user mobile terminal, and can improve resident electricity consumption characteristic analysis efficiency and quickly respond to resident user requirements.

The embodiment of the invention discloses a park management system and a method. The park management system comprises a terminal layer, a transmission layer, a platform layer, an application layer anda display layer, wherein the terminal layer is used for collecting park data information through hardware equipment and/or a third-party subsystem and sending the park data information to the transmission layer; the transmission layer is used for carrying out edge analysis on hardware data information of a set type contained in the park data information and sending the park data information and/or an edge analysis result to the platform layer; the platform layer is used for carrying out data processing and analysis on the park data information and/or the edge analysis result to obtain application data information, and sending the application data information to the application layer; and the application layer is used for issuing a corresponding control instruction to the corresponding hardware equipment and/or the third-party subsystem according to the application data information, so that unified management and control of the hardware equipment and the third-party subsystem containedin the park are realized, and the park management cost is reduced.

The invention discloses an intelligent sensing inspection system and method for a transformer substation, and relates to the technical field of electric power system inspection. The system comprises an edge end data acquisition and analysis subsystem, a station end management subsystem and a cloud artificial intelligence subsystem, the edge end data acquisition and analysis subsystem is used for acquiring target inspection data based on an inspection task and performing edge analysis on the target inspection data based on an artificial intelligence model to obtain an inspection analysis result; the station end management subsystem is used for generating and issuing an inspection task, managing inspection related data and generating an original sample set for training and optimizing an artificial intelligence model; and the cloud artificial intelligence subsystem is used for training and optimizing an artificial intelligence model according to the original sample set and issuing the artificial intelligence model. By multiplexing existing equipment of the transformer substation and deploying intelligent equipment, intelligent and automatic inspection of the transformer substation is realized, the manual operation and maintenance cost is saved, and high-quality operation and maintenance are realized.

The invention discloses a power distribution network digital twin terminal and system and a terminal control method, the power distribution network digital twin terminal comprises a twin model module, a simulation analysis module, a state estimation module and a rehearsal decision module, the twin model module constructs a regional power grid digital twin model, and the simulation analysis model performs power grid operation simulation. The state estimation module evaluates the state of the regional power grid according to the real-time data of the power grid and the simulation result of the simulation analysis model, updates the topology model and the parameter model of the regional power grid according to the state estimation result, and feeds back the updating result to the twin model module for synchronous updating; and the rehearsal decision-making module performs edge analysis decision-making according to the topological model and the parameter model of the regional power grid and the simulation result, and searches and outputs a control strategy aiming at the simulation result from a decision-making library. The technical problems that an existing power grid digital twinning technology has high requirements for the manufacturing cost of a server and is not suitable for a small regional power grid are solved.

The present invention discloses a motor equipment full life-cycle management data acquisition system. The system comprises three layers of structures, the first layer is an acquisition card, the second layer is a data upload card, and the third layer is a data communication card. The acquisition card comprises multiple acquisition card units, and each acquisition card unit acquires the current, the temperature and the vibration signal of one single motor. After the data is screened intelligently, the data is sent to the second-layer data upload card. The data upload card comprises multiple data upload card units, the multiple data upload card units divide the acquisition card units into multiple groups, and each data upload card unit collects the data of one group of acquisition card unitsand sends the data to the third-layer data communication card. The data communication card collects the data of all data upload card units and sends the data to a PC terminal host computer. Accordingto the system, under the complex on-site working condition, an unmanned and intelligent motor current, temperature and vibration acquisition system is established, the data is acquired via various sensors, and the acquired data is subjected to edge analysis, so that the upload bandwidth is reduced to relieve the burden of a communication network, and finally the data is transmitted to a host computer engineer station via a upload plate to perform cloud platform analysis.

The invention discloses a glass edge detection method and system based on machine vision, and the method comprises the steps: carrying out the extraction of a glass edge contour, carrying out the curvature smoothing processing of the extracted contour points, and carrying out the curvature upsampling processing, and respectively fitting the outer contour of the glass edge, defect detection for analyzing forms such as edge breakage, salient points and sawtooth cutting on the edge of the glass is realized. According to the method, the edge of the glass is segmented and extracted based on machine vision, and the defects of chipping corners, sawtooth edges, salient points and the like on the edge of the glass are quickly and stably found out through an edge analysis algorithm. The detection precision is high, the dimension specification is controllable, the labor cost is reduced, the performance is excellent, and the use is convenient.

A method for processing vehicle test data of a vehicle, in which a cloud-based storage and computing device that is communicatively connected to the vehicle via a data network is provided with a multiplicity of modelling vehicle test data that are provided by a plurality of control devices and/or sensor devices of the vehicle. The cloud-based storage and computing device analyzes the received modelling vehicle test data and takes these modelling vehicle test data as a basis for generating test-case-specific edge analytics algorithms in an automated manner and transmits the algorithms to an edge analytics computing device inside the vehicle via the data network. The edge analytics computing device receives a multiplicity of vehicle test data from the control devices and/or the sensor devices during the performance of test or trial drives by the vehicle and evaluates the data by way of the edge analytics algorithms.

The invention relates to the technical field of bionic substations, and discloses a bionic substation based on digitization and intellectualization, which comprises an integrated operation and maintenance management and control platform, the integrated operation and maintenance management and control platform comprises service applications, an NARI-AI algorithm application unit and data collection, the service applications comprise edge service applications and edge internet-of-things agents, and the edge service applications comprise edge service applications and edge internet-of-things agents. The edge service application comprises intelligent inspection, alarm setting, operation and maintenance management, personnel management, comprehensive monitoring and authority management, and the edge internet-of-things agent comprises edge analysis, cloud edge collaboration, protocol adaptation, safety authentication, resource monitoring and container management. According to the invention, the on-site workload is reduced, the real-time on-line anti-misoperation function in a full-service scene is realized, the safety of operation, overhaul, operation and maintenance and other work in the transformer substation is greatly improved, the alarm information containing the accurate position of a fault point is pushed at the first time, precious time is won for maintenance personnel to carry out fault repair work, and the efficiency of the overhaul work is greatly improved.

The invention provides a cerebral apoplexy recurrence risk perception and behavior decision model construction system and method. The risk perception system comprises a data sensing layer, an edge analysis layer, a cloud early warning layer and a risk assessment layer; the data sensing layer sends the physiological measurement parameters to the edge analysis layer; the edge analysis layer executes edge calculation analysis; when the analysis result meets a first preset condition, a feedback signal is sent to the data sensing layer, and the edge calculation analysis results are stored in groups; the cloud early warning layer comprises a plurality of different types of cloud early warning databases; when the analysis result meets a second predetermined condition, the cloud early warning layer executes cloud query calculation in a cloud early warning database based on the edge calculation analysis result stored in groups; and the risk assessment layer receives a cloud query calculation result of the cloud early warning layer and gives a risk assessment value. The invention also discloses a recurrence risk decision system and method. According to the technical scheme, the recurrence risk of the cerebral apoplexy can be effectively perceived.

The invention discloses an edge analysis control method and a regional cooling and heating control system. The regional cooling and heating control system comprises an edge analysis controller, a plurality of devices and a plurality of sensors, wherein the edge analysis controller is used for collecting operation data of the regional cooling and heating control system, analyzing the operation datato obtain an analysis result, and controlling the devices in the regional cooling and heating control system based on the analysis result, specifically, the operation data comprises operation parameters of the devices in the regional cooling and heating control system and environment parameters collected by the sensors in the regional cooling and heating control system. The technical problem thatthe working efficiency of an existing regional cooling and heating control system is low is solved.