Management methods, systems, equipment, and storage media for electricity metering assets in inventory.

Abstract

This invention discloses a method, system, device, and storage medium for managing the inventory of electrical energy metering assets, relating to the field of power material warehousing management technology. The method includes: responding to a trigger command for acquiring an image of an electrical energy metering asset, activating a servo motor to drive a base on a shelf to move, enabling an image acquisition device on the base to acquire an image of the electrical energy metering asset; inputting the image into a pre-trained target detection model to obtain the bounding boxes and category confidence scores of each electrical energy metering asset in the image; extracting barcodes within each bounding box based on the bounding boxes and a pre-trained barcode detection model; determining the center coordinates of each bounding box based on its position and a pre-calibrated coordinate reference; generating a first asset list based on the center coordinates of each bounding box, the corresponding barcode, and the category confidence score; and comparing the first asset list with a second asset list to determine the inventory status of each electrical energy metering asset. This improves the efficiency of determining the inventory status of assets.

Description

Technical Field

[0001] This invention relates to the field of power material storage management technology, specifically to a method, system, equipment, and storage medium for managing in-warehouse power metering assets. Background Technology

[0002] Electricity metering assets refer to the specialized key equipment used by power grid companies for the transmission of electricity value, electricity metering, and collection of electricity consumption information. They are important fixed assets and metering instruments of power companies.

[0003] The standardization of the management of electricity metering assets directly affects the accuracy of metering, asset security, and compliant operation. Currently, grassroots warehouse sites generally use manual registration or semi-automatic scanning methods to manage the entry and exit of electricity metering assets. This is cumbersome, records are delayed, and it is difficult to efficiently ascertain the status of assets in the warehouse. Summary of the Invention

[0004] The purpose of this invention is to provide a method, system, device, and storage medium for managing the inventory of electrical energy metering assets. By collecting and comparing images of the inventory of electrical energy metering assets, the inventory status of the assets can be quickly determined, thus solving the problem of the difficulty in efficiently knowing the inventory status of assets.

[0005] This invention is achieved through the following technical solution:

[0006] The first aspect of this application provides a method for managing the inventory of electricity metering assets, including:

[0007] In response to the trigger command for acquiring images of electricity metering assets, a servo motor is started to drive the base on the shelf to move, so that the image acquisition device installed on the base can acquire images of all electricity metering assets placed on the shelf as the base moves.

[0008] The image is input into a pre-trained object detection model to obtain the bounding boxes of each electricity metering asset in the image and the category confidence of each electricity metering asset output by the object detection model.

[0009] Based on the bounding boxes and the pre-trained barcode detection model, the barcodes within each bounding box are extracted;

[0010] Based on the position of the bounding boxes and the pre-calibrated coordinate reference, determine the center coordinates of each bounding box;

[0011] A first asset list is generated based on the center coordinates of each bounding box, the corresponding barcode, and the category confidence level.

[0012] The first asset list is compared with the second asset list generated based on the previous image acquisition results, and the inventory status of each electricity metering asset is determined according to the comparison results; the inventory status includes new entry, exit, asset movement and abnormal status.

[0013] In one feasible implementation, the acquisition trigger instruction is triggered based on at least one of the following behaviors:

[0014] The visual recognition equipment installed in the asset warehouse detects human activity in the shelf area, reaches the predetermined image acquisition time, and is manually initiated for image acquisition.

[0015] In one feasible implementation, the start servo motor drives the base on the shelf to move, causing the image acquisition device mounted on the base to acquire images of all electrical energy metering assets placed on the shelf as the base moves, including:

[0016] A servo motor is activated to drive the base on the shelf to move along a linear track, so that the image acquisition device installed on the base can acquire an image of a portion of the electrical energy metering assets on the shelf when it reaches a predetermined position; the predetermined position includes at least two pre-set positions.

[0017] The control base moves one revolution along the linear track, and the images of the acquired electrical energy metering assets are stitched together to obtain an image of all the electrical energy metering assets placed on the shelf.

[0018] In one feasible implementation, the abnormal state is determined in the following way:

[0019] If a new valid barcode that does not exist in the second asset list appears in the first asset list, and the confidence level of the category corresponding to the new valid barcode in the first asset list is less than a preset value, then the image of the location corresponding to the new valid barcode is re-acquired, and a third asset list is formed.

[0020] If the confidence level of the category corresponding to the new valid barcode in the third asset list is still less than the preset value, it is determined that there is an abnormal state. The specific type of the abnormal state is that illegal assets with valid barcodes are placed on the shelf.

[0021] In one feasible implementation, if the database status is abnormal and the abnormality type cannot be identified, the method further includes:

[0022] The image is uploaded to a cloud processor, which then performs the following: inputting the image into a pre-trained anomaly detection model to output anomaly detection results; the anomaly detection results include anomaly location, anomaly type, and anomaly confidence; the anomaly type includes at least one of the following: equipment damage, missing parts, illegal assets with valid barcodes, and assets with invalid barcodes.

[0023] Receive anomaly detection results from the cloud;

[0024] The anomaly detection model is trained using a dataset of anomaly images uploaded from various sites.

[0025] In one feasible implementation, the method further includes:

[0026] In response to a positioning command for the target asset, the target center coordinates of the target asset are obtained from a pre-set database;

[0027] The servo motor is activated to drive the base to move to a predetermined position closest to the center coordinates of the target; the predetermined position is a pre-set position for image acquisition.

[0028] Based on the coordinates of the predetermined position, the coordinates of the gimbal laser on the base, and the coordinates of the target center, the rotation angle of the gimbal laser is determined.

[0029] The gimbal laser is controlled to emit laser light according to the rotation angle to illuminate the target asset.

[0030] In one feasible implementation, the shelf includes at least one shelf on which different types of the electrical energy metering assets are placed.

[0031] A second aspect of this application provides a management system for storing electricity metering assets, including:

[0032] The image acquisition unit, in response to the acquisition trigger command of the image of the electricity metering asset, starts the servo motor to drive the base on the shelf to move, so that the image acquisition device installed on the base can acquire images of all the electricity metering assets placed on the shelf as the base moves.

[0033] The object detection unit is used to input the image into a pre-trained object detection model to obtain the bounding boxes of each electricity metering asset in the image and the category confidence of each electricity metering asset output by the object detection model.

[0034] The barcode extraction unit extracts the barcodes within each bounding box based on the bounding boxes and the pre-trained barcode detection model.

[0035] The center coordinate determination unit determines the center coordinates of each bounding box based on the position of the bounding box and the pre-calibrated coordinate reference.

[0036] The asset list generation unit generates the first asset list based on the center coordinates of each bounding box, the corresponding barcode, and the category confidence level.

[0037] The asset status determination unit is used to compare the first asset list with the second asset list generated based on the previous image acquisition results, and determine the inventory status of each electricity metering asset according to the comparison results; the inventory status includes new entry, exit, asset movement and abnormal status.

[0038] A third aspect of this application provides an electronic device, including: a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the above-described method.

[0039] A fourth aspect of this application provides a storage medium, comprising: storing a program or instructions on the storage medium, wherein the program or instructions, when executed by a processor, implement the steps of the above-described method.

[0040] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0041] This application embodiment acquires images of electricity metering assets on shelves, performs target detection on the images to identify the bounding boxes and category confidence levels of each electricity metering asset in the images, and identifies the barcodes within the bounding boxes to form a (first) asset list corresponding to the image; by comparing the asset lists formed by the two image acquisitions, the presence of valid barcodes in the two asset lists, the displacement of the center coordinates of the asset bounding boxes, and the value of the category confidence level are used to determine the status of each electricity metering asset, such as new arrival, departure, asset movement, and abnormality; no manual review of records or on-site inventory is required, which improves the efficiency of knowing the status of assets in the warehouse. Attached Figure Description

[0042] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:

[0043] Figure 1 A flowchart illustrating a method for managing the inventory of electrical energy metering assets provided in this application embodiment;

[0044] Figure 2 A schematic diagram of the structure of the shelving used at the site in a method for managing the inventory of electrical energy metering assets provided in this application embodiment;

[0045] Figure 3 A schematic diagram of the structure of a management system for storing electricity metering assets is provided in an embodiment of this application;

[0046] Figure 4 This is a schematic diagram of the structure of a computing device provided in an embodiment of this application;

[0047] In the diagram, 21 is the shelf body, 22 is the linear guide rail, 23 is the base, 24 is the industrial camera, 25 is the pan-tilt laser light, 26 is the shelf, 31 is the image acquisition unit, 32 is the target detection unit, 33 is the barcode extraction unit, 34 is the center coordinate determination unit, 35 is the asset list generation unit, 36 is the asset status determination unit, 41 is the memory, 42 is the processor, 43 is the communication component, 44 is the display, 45 is the power supply component, and 46 is the audio component. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments and accompanying drawings. The illustrative embodiments and descriptions of this invention are for explanation only and are not intended to limit the invention. All other embodiments obtained by those skilled in the art based on the embodiments in this application without creative effort are within the scope of protection of this application.

[0049] As will be known to those skilled in the art, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0050] The terms “comprising” and “having”, and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not expressly listed or inherent to such process, method, product, or apparatus.

[0051] Example 1:

[0052] Embodiment 1 of this application provides a method for managing the inventory of electricity metering assets, in order to solve the problem of difficulty in efficiently knowing the inventory status of assets.

[0053] The subject executing this method can be any computing device capable of implementing the method, such as a server, mobile phone, personal computer, smart wearable device, smart robot, etc.

[0054] Furthermore, the embodiments of this application do not limit the execution order of different steps. When using the method provided in the embodiments of this application, the execution order of different steps can be adjusted according to actual needs.

[0055] For ease of description, the following uses a management device for managing electricity metering assets as the subject of this method to provide a detailed description of the method provided in this application embodiment.

[0056] like Figure 1 The diagram shown is a flowchart illustrating the specific implementation of a method for managing the inventory of electricity metering assets according to an embodiment of this application, including the following steps 11-16:

[0057] Step 11: In response to the acquisition trigger command for the image of the electricity metering assets, start the servo motor to drive the base on the shelf to move, so that the image acquisition device installed on the base can acquire images of all the electricity metering assets placed on the shelf as the base moves.

[0058] Electricity metering assets can be of different types, such as electricity meters, low-voltage current transformers, data acquisition terminals, communication modules, and other specialized equipment used for the transmission and measurement of electricity values.

[0059] The acquisition trigger command, i.e., the command that triggers image acquisition, can be triggered based on at least one of the following behaviors:

[0060] The visual recognition equipment installed in the asset warehouse detects human activity in the shelf area, reaches the predetermined image acquisition time, and is manually initiated for image acquisition.

[0061] The visual acquisition device can be a wide-angle visual recognition module installed at the center of the top of the shelf, capable of detecting personnel activity within the shelf area. To detect all locations within the asset warehouse, multiple wide-angle visual recognition modules can be installed. Specifically, this wide-angle visual recognition module can be a wide-angle industrial camera, industrial vision module, or other device with camera and image analysis capabilities.

[0062] In this embodiment, the visual acquisition module continuously acquires and analyzes the video stream of the shelf area. After detecting human activity in the video stream, it sends an acquisition trigger command to the management device of this embodiment.

[0063] In order to achieve real-time control over the status of the electrical energy metering assets in the warehouse, the management device in this embodiment will pre-set the image acquisition time, such as acquiring images once every 1 hour, and sending an acquisition trigger command after the predetermined time is reached.

[0064] In specific situations, such as when managers want to know the inventory status of electricity metering assets in real time, they can directly send a data collection trigger command.

[0065] like Figure 2 As shown, the shelf in this embodiment includes a shelf body 21 and at least one shelf 26 horizontally arranged on the shelf body 21. The shelf 26 is used to place electrical energy metering assets. Except for the bottom shelf, each shelf is provided with a linear guide rail 22 below it. A base 23 is provided on the linear guide rail 22. An image acquisition device—an industrial camera 24—and a pan-tilt laser light 25 are installed on the base 23.

[0066] The base 23 is driven by a servo motor and moves on the linear guide rail 22. The servo motor, industrial camera 24, gimbal laser light 25, and vision recognition equipment are connected to the industrial control host, which supports data synchronization with terminals (mobile APP or PC client). The industrial control host can be the management device in this embodiment.

[0067] Driven by the base, the industrial camera captures images of the power metering assets at different locations on the lower shelf; the pan-tilt laser light uses laser illumination to locate the target power metering assets on the lower shelf.

[0068] In this step, a servo motor is activated to drive the base on the shelf to move, causing the image acquisition device mounted on the base to acquire images of all the electrical energy metering assets placed on the shelf as the base moves, including:

[0069] The servo motor is started to drive the base on the shelf to move along a linear track, so that the image acquisition device installed on the base can acquire images of some of the power metering assets on the shelf when it reaches a predetermined position; the predetermined position includes at least two pre-set positions; the base is controlled to move one revolution along the linear track, and the images of the acquired power metering assets are stitched together to obtain an image of all the power metering assets placed on the shelf.

[0070] In this process, N (where N is an integer not less than 2) predetermined positions are set on the linear track, dividing the linear track into N+1 segments. The selection of N must satisfy the following condition: when the base drives the industrial camera to move one revolution on the linear track, the field of view of the industrial camera must cover the lower shelf space. The predetermined positions can be determined based on the pulse count of the servo motor. For example, if the servo motor drives the base to move at a constant speed on the linear track, the mapping relationship between the motor pulse count and each predetermined position is calibrated. When the target motor pulse count is reached, the corresponding predetermined position in the mapping relationship is reached, and the industrial camera is controlled to perform the taking action.

[0071] The images of the various electrical energy metering assets can be stitched together using a computer vision processing model to obtain images of all electrical energy metering assets on the lower shelf of the industrial camera. The computer vision processing model used can be one of the following algorithms from the OpenCV tool library: SIFT (Scale-Invariant Feature Transform), SURT (Speeded-Up Robust Features, a computer vision algorithm for image feature detection and description), or ORB (Oriented FAST and Rotated BRIEF). This involves extracting stable keypoints and feature descriptions from multiple adjacent partial images; calculating the similarity of features between adjacent images to find matching point pairs; eliminating mismatches using the RANSAC (Random Sample Consensus) algorithm; calculating the projection transformation relationship between images (homography matrix H); projecting all local images onto a unified panoramic coordinate system based on the homography matrix; and performing weighted fusion or Laplacian fusion on overlapping areas to eliminate stitching seams and generate a complete panoramic image—an image of all electrical metering assets on a single shelf.

[0072] To reduce storage space waste caused by changes in the proportion of different equipment types, maximize shelf utilization efficiency, and reduce investment and resource requirements, different types of electricity metering assets can be flexibly placed on the same shelf 26 to achieve storage space sharing. Additionally, electricity metering assets are typically not placed on the top shelf.

[0073] Step 12: Input the image into a pre-trained target detection model to obtain the bounding boxes of each electricity metering asset in the image and the category confidence of each electricity metering asset output by the target detection model.

[0074] The object detection model in this embodiment can be an artificial intelligence model that integrates object detection and image segmentation, such as an object detection model based on the YOLO series or MMDetection framework. This object detection model uses multi-angle image samples and partially occluded image samples of various types of electricity metering assets as datasets. Boundary box annotations and category annotations are used as labels to train the object detection model. The training method of traditional deep neural network models is employed to obtain the trained object detection model.

[0075] The category confidence score output by the object detection model is used to distinguish whether the object within the identified bounding box belongs to a certain type of electricity metering asset or an illegal asset (foreign object).

[0076] Step 13: Based on the bounding boxes and the pre-trained barcode detection model, extract the barcodes within each bounding box.

[0077] The barcode detection model used in this embodiment can be a traditional OpenCV barcode recognition model. The process of extracting barcodes using this barcode recognition model can be as follows: by performing grayscale conversion, Gaussian denoising, edge / gradient enhancement, and binarization on each bounding box identified in step 12, filtering rectangular strip outlines (by aspect ratio, area, and rectangularity), straightening tilted / curved barcodes to a standard frontal view, enhancing the alternating black and white bar texture, calling the ZBar / ZXing kernel, and outputting the barcode number; the barcode number is then identified to determine the valid barcode used to identify electricity metering assets.

[0078] Step 14: Determine the center coordinates of each bounding box based on its position and the pre-calibrated coordinate reference.

[0079] The coordinate reference can be calibrated by attaching a visual calibration board (such as an ArUco calibration board, 5×5 grids, each grid 30mm×30mm, with an error of ±0.05mm) to a designated corner of each shelf. This is used to calibrate the origin of the stitched image coordinates. By comparing the pixel coordinates of the asset with those of the calibration board, the pixel coordinates of the asset on the shelf (such as the center coordinates) can be determined. The calibration board ID can also serve as a unique identifier for the shelf.

[0080] By using the location of the bounding boxes and the pre-calibrated coordinate references, the coordinates of each bounding box can be determined, thereby further determining the center coordinates of each bounding box.

[0081] Step 15: Generate the first asset list based on the center coordinates of each bounding box, the corresponding barcode, and the category confidence level.

[0082] The first asset list contains the center coordinates of each bounding box, and the mapping relationship between its corresponding barcode and category confidence level.

[0083] Step 16: Compare the first asset list with the second asset list generated based on the previous image acquisition results, and determine the inventory status of each electricity metering asset according to the comparison results; the inventory status includes new entry, exit, asset movement and abnormal status.

[0084] By comparing the first and second asset lists, and based on the presence of valid barcodes in the two asset lists, the displacement of the center coordinates of the asset boundary boxes, and the category confidence level, the status of each electricity metering asset, such as new entry, exit, asset movement, and abnormality, is determined.

[0085] Specifically:

[0086] If a new valid barcode appears in the first asset list that is not present in the second asset list, and the category confidence level corresponding to the new valid barcode exceeds a preset value (e.g., 0.85), it is considered that a new electricity metering asset has been added to the database. The bounding box center coordinates of the new electricity metering asset, along with its corresponding asset barcode and category confidence level information, are stored in the database.

[0087] If the first asset list is missing a valid barcode that exists in the second asset list, it is assumed that the electricity metering asset corresponding to the missing valid barcode has been removed from the warehouse. The bounding box center coordinates of the relevant electricity metering asset, as well as its corresponding asset barcode and category confidence information, are deleted from the database.

[0088] If a valid barcode exists in both lists and the corresponding category confidence is higher than the preset value, but its corresponding center coordinates are different in the two lists, it is considered that the target asset corresponding to the valid barcode has been displaced, and the center coordinates of the asset bounding box are updated in the database.

[0089] If a new valid barcode appears in the first asset list that is not present in the second asset list, and the confidence level of the category corresponding to the new valid barcode in the first asset list is less than a preset value, then the image of the location corresponding to the new valid barcode is re-acquired, and a third asset list is formed. If the confidence level of the category corresponding to the new valid barcode in the third asset list is still less than the preset value, then an abnormal state is determined to exist, and the specific type of the abnormal state is that illegal assets with valid barcodes are placed on the shelf. By identifying this type of abnormal state, it is possible to prevent relevant personnel from falsifying the asset's inventory status to evade responsibility after the asset is lost.

[0090] Once an illegal asset with a valid barcode is detected, an alarm is triggered, and the person responsible for the asset is notified.

[0091] In one feasible implementation, to further confirm the legality of asset release, this embodiment also includes tracing the target asset corresponding to the missing valid barcode. The specific tracing process may include: tracing whether a gimbal laser light was used to locate the target asset within the interval between two adjacent image acquisitions; if so, the target asset is considered to have been legally released; and tracing whether an authorized staff member entered the shelf area within the interval between two adjacent image acquisitions; if so, checking the staff member's permissions; if the staff member has the permission to access the target asset, the target asset is considered to have been legally released; in any other case, the target asset is considered to have been illegally released.

[0092] In one feasible implementation, if the inventory status at the site where the shelf is placed is determined to be abnormal, and the type of abnormality cannot be identified, this embodiment further includes:

[0093] The image is uploaded to a cloud processor, which then performs the following actions: inputting the image into a pre-trained anomaly detection model to output anomaly detection results; the anomaly detection results include anomaly location, anomaly type, and anomaly confidence level; the anomaly type includes at least one of equipment damage, missing parts, illegal assets with valid barcodes, and assets with invalid barcodes; receiving the anomaly detection results output from the cloud and sending them to the asset manager. The asset manager verifies the received anomaly detection results and feeds the verification results back to the cloud processor as samples for incremental training of the cloud anomaly detection model.

[0094] Anomaly detection models can be high-precision anomaly detection models, such as YOLOv8-X, YOLOv10-X (ultra-precision version), Cascade R-CNN, etc. The cloud-based anomaly detection model is trained on a dataset of anomaly images uploaded from various sites. This dataset includes at least images of illegitimate assets with valid barcodes, images of legitimate assets with damaged or missing casings, images of legitimate assets with missing accessories such as watch covers and / or terminals, as well as anomaly images such as foreign objects, fake assets, and blurred / counterfeit barcodes.

[0095] In this implementation, a closed-loop logic for anomaly assessment is constructed through cloud-edge collaboration. Distributed shelf edge detection data and feedback from responsible parties are aggregated to drive incremental training of the cloud model and dynamic threshold optimization for anomaly judgment, thereby achieving continuous self-evolution of the accuracy of anomaly alarms and the robustness of the system.

[0096] In one feasible implementation, to more quickly locate the target asset, this embodiment also includes:

[0097] In response to a positioning command for the target asset, the target center coordinates of the target asset are obtained from a pre-set database;

[0098] The servo motor is activated to drive the base to move to a predetermined position closest to the center coordinates of the target; the predetermined position is a pre-set position for image acquisition.

[0099] Based on the coordinates of the predetermined position, the coordinates of the gimbal laser on the base, and the coordinates of the target center, the rotation angle of the gimbal laser is determined.

[0100] The gimbal laser is controlled to emit laser light according to the rotation angle to illuminate the target asset.

[0101] The target asset location command can be triggered when executing a related work order (e.g., executing a work order to retrieve the target asset) or when viewing the target asset's location in the warehouse.

[0102] It should be noted that each electricity metering asset has a unique identifier, which can be represented by a valid barcode affixed to the asset. In this implementation, the corresponding target center coordinates can be determined by searching for the valid barcode of the target asset in the database.

[0103] The predetermined position in this implementation can be the position marked when acquiring images.

[0104] Based on the coordinates of the predetermined position, the coordinates of the gimbal laser on the base, and the coordinates of the target center, the rotation angle of the gimbal laser is determined, specifically:

[0105] Based on the target center coordinates ( , ,0), the coordinates of the nearest predetermined location ( (0, 0), and the coordinates of the gimbal laser light when the base moves to the predetermined position ( , , Then, the distance vector l from the gimbal laser light to the target asset can be expressed as: ( ,0,0)-( , ,0); calculate the rotation angle (θ, φ) of the gimbal laser light based on the distance vector l, θ=arctan( / ), φ=arctan( ).

[0106] In summary, this application embodiment acquires images of electricity metering assets on shelves, performs target detection on the images to identify the bounding boxes and category confidence levels of each electricity metering asset in the images, and identifies the barcodes within the bounding boxes to form a (first) asset list corresponding to the image; by comparing the asset lists formed from the two image acquisitions, the presence of valid barcodes in the two asset lists, the displacement of the center coordinates of the asset bounding boxes, and the category confidence levels in the comparison results, the status of each electricity metering asset, such as new entry, exit, asset movement, and abnormality, is determined; no manual review of records or on-site inventory is required, thus improving the efficiency of knowing the status of assets in storage.

[0107] Example 2:

[0108] To address the problem of difficulty in efficiently obtaining the status of assets in the existing technology, and based on the same inventive concept as Embodiment 1, this application also provides a management system for the inventory of electricity metering assets.

[0109] The specific structural diagram of the system is as follows: Figure 3 As shown, it includes the following functional units:

[0110] Image acquisition unit 31, in response to the acquisition trigger command of the image of the power metering asset, starts the servo motor to drive the base on the shelf to move, so that the image acquisition device installed on the base can acquire images of all the power metering assets placed on the shelf as the base moves.

[0111] The target detection unit 32 is used to input the image into a pre-trained target detection model to obtain the bounding boxes of each energy metering asset in the image and the category confidence of each energy metering asset output by the target detection model.

[0112] The barcode extraction unit 33 extracts the barcodes within each bounding box based on the bounding boxes and the pre-trained barcode detection model.

[0113] The center coordinate determination unit 34 determines the center coordinates of each bounding box based on the position of the bounding box and the pre-calibrated coordinate reference.

[0114] The asset list generation unit 35 generates the first asset list based on the center coordinates of each bounding box, the corresponding barcode, and the category confidence level.

[0115] The asset status determination unit 36 ​​is used to compare the first asset list with the second asset list generated based on the previous image acquisition results, and determine the in-stock status of each electricity metering asset according to the comparison results; the in-stock status includes new entry, exit, asset movement and abnormal status.

[0116] The acquisition trigger command is based on at least one of the following behaviors:

[0117] The visual recognition equipment installed in the asset warehouse detects human activity in the shelf area, reaches the predetermined image acquisition time, and is manually initiated for image acquisition.

[0118] The image acquisition unit 31 is specifically used to: start the servo motor to drive the base on the shelf to move along the linear track, so that the image acquisition device installed on the base can acquire images of some of the power metering assets on the shelf when it reaches a predetermined position; the predetermined position includes at least two pre-set positions; control the base to move one revolution along the linear track, and stitch together the acquired images of each part of the power metering assets to obtain an image of all the power metering assets placed on the shelf.

[0119] The abnormal status determined by asset status determination unit 36 ​​is determined in the following ways:

[0120] If a new valid barcode appears in the first asset list that is not present in the second asset list, and the confidence level of the category corresponding to the new valid barcode in the first asset list is less than a preset value, then the image of the location corresponding to the new valid barcode is re-acquired, and a third asset list is formed; if the confidence level of the category corresponding to the new valid barcode in the third asset list is still less than the preset value, then an abnormal state is determined to exist, and the type of the abnormal state is specifically that illegal assets with valid barcodes are placed on the shelf.

[0121] The management system in this embodiment also includes a cloud-edge interaction unit, specifically used for:

[0122] When the image is in an abnormal state and the type of abnormality cannot be identified, the image is uploaded to the cloud processor, which then performs the following actions: inputting the image into a pre-trained anomaly detection model to output anomaly detection results; the anomaly detection results include anomaly location, anomaly type, and anomaly confidence; the anomaly type includes at least one of equipment damage, missing parts, illegal assets with valid barcodes, and assets with invalid barcodes; and receiving the anomaly detection results output from the cloud.

[0123] The anomaly detection model was trained using a dataset of anomaly images uploaded from various sites.

[0124] The management system in this embodiment also includes a target asset location unit, specifically used for:

[0125] In response to a positioning command for the target asset, the target center coordinates of the target asset are obtained from a pre-set database; a servo motor is activated to drive the base to move to a predetermined position closest to the target center coordinates; the predetermined position is a pre-set position for image acquisition; based on the coordinates of the predetermined position, the coordinates of the gimbal laser on the base, and the target center coordinates, the rotation angle of the gimbal laser is determined; the gimbal laser is controlled to emit laser light according to the rotation angle to illuminate the target asset.

[0126] In addition, the shelf in this embodiment includes at least one shelf, on which different types of electrical energy metering assets are placed.

[0127] This application embodiment acquires images of electricity metering assets on shelves, performs target detection on the images to identify the bounding boxes and category confidence levels of each electricity metering asset in the images, and identifies the barcodes within the bounding boxes to form a (first) asset list corresponding to the image; by comparing the asset lists formed by the two image acquisitions, the presence of valid barcodes in the two asset lists, the displacement of the center coordinates of the asset bounding boxes, and the category confidence levels in the comparison results, the status of each electricity metering asset, such as new entry, exit, asset movement, and abnormality, is determined; no manual review of records or on-site inventory is required, which improves the efficiency of knowing the status of assets in the warehouse.

[0128] Based on the same inventive concept as the foregoing embodiments of this application, this application also provides a computing device.

[0129] like Figure 4 As shown, the computing device includes a memory 41 and a processor 42. The memory 41 can be configured to store various other data to support operation on the electronic device. Examples of such data include instructions for any application or method used to operate on the electronic device. The memory 41 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0130] The processor 42, coupled to the memory 41, is used to execute the computer program stored in the memory 41 to perform a method for managing the inventory of electrical energy metering assets as described in the foregoing embodiments.

[0131] When the processor 42 executes the computer program to perform the management method for the inventory of electrical energy metering assets, it acquires images of electrical energy metering assets on the shelves, performs target detection on the images to identify the bounding boxes and category confidence levels of each electrical energy metering asset in the images, and identifies the barcodes within the bounding boxes to form a (first) asset list corresponding to the image. By comparing the asset lists formed by the two image acquisitions, the presence of valid barcodes in the two asset lists, the displacement of the center coordinates of the asset bounding boxes, and the category confidence levels in the comparison results, the status of each electrical energy metering asset, such as new entry, exit, asset movement, and abnormality, is determined. This eliminates the need for manual review of records or on-site inventory, improving the efficiency of knowing the inventory status of assets.

[0132] When the processor 42 executes the computer program in the memory 41, in addition to the functions described above, it can also perform other functions, as detailed in the descriptions of the preceding embodiments.

[0133] Furthermore, such as Figure 4 As shown, the computing device also includes other components such as a display 44, a communication component 43, a power supply component 45, and an audio component 46. Figure 4 The diagram only shows some components and does not mean that the computing device includes only these components. Figure 4 The components shown.

[0134] Accordingly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a computer, can implement the methods provided in the above embodiments.

[0135] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0136] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments.

[0137] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for managing the inventory of electricity metering assets, characterized in that, include: In response to the trigger command for acquiring images of electricity metering assets, a servo motor is started to drive the base on the shelf to move, so that the image acquisition device installed on the base can acquire images of all electricity metering assets placed on the shelf as the base moves. The image is input into a pre-trained object detection model to obtain the bounding boxes of each electricity metering asset in the image and the category confidence of each electricity metering asset output by the object detection model. Based on the bounding boxes and the pre-trained barcode detection model, the barcodes within each bounding box are extracted; Based on the position of the bounding boxes and the pre-calibrated coordinate reference, determine the center coordinates of each bounding box; A first asset list is generated based on the center coordinates of each bounding box, the corresponding barcode, and the category confidence level. The first asset list is compared with the second asset list generated based on the previous image acquisition results, and the inventory status of each electricity metering asset is determined according to the comparison results; the inventory status includes new entry, exit, asset movement and abnormal status.

2. The method for managing the inventory of electricity metering assets according to claim 1, characterized in that, The acquisition trigger command is triggered based on at least one of the following behaviors: The visual recognition equipment installed in the asset warehouse detects human activity in the shelf area, reaches the predetermined image acquisition time, and is manually initiated for image acquisition.

3. The method for managing the inventory of electricity metering assets according to claim 1, characterized in that, The starting servo motor drives the base on the shelf to move, so that the image acquisition device installed on the base can acquire images of all the electricity metering assets placed on the shelf as the base moves, including: A servo motor is activated to drive the base on the shelf to move along a linear track, so that the image acquisition device installed on the base can acquire an image of a portion of the electrical energy metering assets on the shelf when it reaches a predetermined position; the predetermined position includes at least two pre-set positions. The control base moves one revolution along the linear track, and the images of the acquired electrical energy metering assets are stitched together to obtain an image of all the electrical energy metering assets placed on the shelf.

4. The method for managing the inventory of electricity metering assets according to claim 1, characterized in that, The abnormal state is determined in the following way: If a new valid barcode that does not exist in the second asset list appears in the first asset list, and the confidence level of the category corresponding to the new valid barcode in the first asset list is less than a preset value, then the image of the location corresponding to the new valid barcode is re-acquired, and a third asset list is formed. If the confidence level of the category corresponding to the new valid barcode in the third asset list is still less than the preset value, it is determined that there is an abnormal state. The specific type of the abnormal state is that illegal assets with valid barcodes are placed on the shelf.

5. The method for managing the inventory of electricity metering assets according to claim 1, characterized in that, If the in-stock status is abnormal and the abnormality type cannot be identified, the method further includes: The image is uploaded to a cloud processor, which then performs the following: inputting the image into a pre-trained anomaly detection model to output anomaly detection results; the anomaly detection results include anomaly location, anomaly type, and anomaly confidence; the anomaly type includes at least one of the following: equipment damage, missing parts, illegal assets with valid barcodes, and assets with invalid barcodes. Receive anomaly detection results from the cloud; The anomaly detection model is trained using a dataset of anomaly images uploaded from various sites.

6. The method for managing the inventory of electricity metering assets according to claim 1, characterized in that, The method further includes: In response to a positioning command for the target asset, the target center coordinates of the target asset are obtained from a pre-set database; The servo motor is activated to drive the base to move to a predetermined position closest to the center coordinates of the target; the predetermined position is a pre-set position for image acquisition. Based on the coordinates of the predetermined position, the coordinates of the gimbal laser on the base, and the coordinates of the target center, the rotation angle of the gimbal laser is determined. The gimbal laser is controlled to emit laser light according to the rotation angle to illuminate the target asset.

7. The method for managing the inventory of electricity metering assets according to claim 1, characterized in that, The shelf includes at least one shelf, on which different types of the electricity metering assets are placed.

8. A management system for electricity metering assets, characterized in that, include: The image acquisition unit, in response to the acquisition trigger command of the image of the electricity metering asset, starts the servo motor to drive the base on the shelf to move, so that the image acquisition device installed on the base can acquire images of all the electricity metering assets placed on the shelf as the base moves. The object detection unit is used to input the image into a pre-trained object detection model to obtain the bounding boxes of each electricity metering asset in the image and the category confidence of each electricity metering asset output by the object detection model. The barcode extraction unit extracts the barcodes within each bounding box based on the bounding boxes and the pre-trained barcode detection model. The center coordinate determination unit determines the center coordinates of each bounding box based on the position of the bounding box and the pre-calibrated coordinate reference. The asset list generation unit generates the first asset list based on the center coordinates of each bounding box, the corresponding barcode, and the category confidence level. The asset status determination unit is used to compare the first asset list with the second asset list generated based on the previous image acquisition results, and determine the inventory status of each electricity metering asset according to the comparison results; the inventory status includes new entry, exit, asset movement and abnormal status.

9. An electronic device, characterized in that, include: A processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method as claimed in any one of claims 1-7.

10. A storage medium, characterized in that, include: The storage medium stores a program or instructions that, when executed by a processor, implement the steps of the method as described in any one of claims 1-7.

Images