An intelligent identification method, system and medium for distribution workers wearing

Abstract

The present invention discloses a method, system, and medium for intelligently identifying the wearable equipment of distribution workers. The method belongs to the field of big data and distribution operation technology. The method comprises: performing image recognition on workers and their wearable equipment in surveillance videos; performing position detection on worker image information; matching and associating safety factor information with workers based on the position information; performing personnel attribute recognition based on multiple detection results to obtain location area information and personnel attribute information; extracting standard wear requirement data for workers to determine the wearable conditions at trajectory points; and performing threshold judgment based on the wearable conditions of the entire trajectory. Thus, based on big data and image recognition and analysis technology, the wearable equipment of distribution workers is identified, and the wearable conditions of the personnel trajectory are identified and evaluated based on the obtained standard wearable conditions, thereby realizing identification and detection of the wearable equipment of workers, improving effective identification and monitoring of the wearable equipment of workers, and ensuring personnel safety.

Description

Technical Field

[0001] The present invention relates to the technical field of big data and power distribution operations, and in particular to an intelligent identification method, system and medium worn by power distribution operators. Background Art

[0002] Currently, the protection and safety monitoring methods for distribution workers usually adopt manual monitoring and protect people through common equipment and wearables. However, there is a lack of tracking and identification technology for personnel wearable standards. It is impossible to effectively identify and supervise the protective wear of distribution workers of different work types according to the distribution operation environment and location, and it is impossible to implement intelligent technology for identifying, alarming and supervising incorrect wearables during distribution operations.

[0003] Based on video monitoring methods and image recognition technology, an intelligent distribution worker wear identification technology is established to perform image recognition of the distribution operation environment, location, and type of workers, identify standard wear requirements, and make wear judgments and identifications throughout the entire construction process. This is a scenario technology means that is currently not available and is also a technical gap in the field of distribution service technology.

[0004] In response to the above problems, effective technical solutions are urgently needed. Summary of the Invention

[0005] The purpose of the embodiments of the present application is to provide an intelligent identification method, system and medium for distribution workers' wear, which can identify the wear of distribution workers based on big data and image recognition and analysis technology, and identify and evaluate the wear status of personnel trajectories based on the obtained standard wear, realize the identification and detection of workers' wear, improve the effective identification and monitoring of workers' wear, and ensure personnel safety.

[0006] The present application also provides an intelligent identification method for power distribution workers wearing clothing, including the following steps:

[0007] Collect real-time monitoring videos of workers working on the distribution network without power outages, perform image recognition on the workers and their equipment in the monitoring videos, and obtain worker image information, equipment information, and safety factor information of the working environment;

[0008] The position of the operator image information is detected by the established multi-scale human body detection model, the safety factor information is supplemented according to the detected position information, and matched with the operator, and the operator, position and safety factor information are mapped and associated;

[0009] Based on multiple detection results obtained through multiple matching and association identification, information is integrated and person attributes are identified to obtain the location area information and person attribute information to be tracked;

[0010] Extracting standard wearing requirement data of the operator through the distribution network operation wear database according to the location area information to be tracked and the operator attribute information;

[0011] Determine the target worker's trajectory information based on the intersection-and-union ratio of the target worker's tracking frame at each trajectory point. If the intersection-and-union ratio of a trajectory point is less than a preset threshold, it is determined that the target worker monitored at the trajectory point does not exist, and the trajectory point is removed.

[0012] Recording the valid track points in the target worker's track information and the corresponding wearing information, and determining the wearing condition of the target worker at each valid track point according to the standard wearing requirement data;

[0013] A threshold judgment is made based on the full-trajectory wearing status of each operator within the preset monitoring time period, and operators who exceed the wearing warning threshold are warned and recorded.

[0014] Optionally, in the intelligent identification method for distribution workers' wearables described in an embodiment of the present application, the method collects real-time monitoring videos of distribution network workers without power outages, performs image recognition on the workers and their wearables in the monitoring videos, and obtains worker image information, wearable information, and safety factor information of the working environment, including:

[0015] Collect real-time monitoring video of the non-stop operation area of ​​the distribution network within the preset area;

[0016] Acquire images of workers in real-time surveillance videos, perform image recognition and image information extraction on workers and wearable devices, and obtain worker image information and wearable device information;

[0017] Based on image recognition, safety factor information of the power distribution operation environment is extracted, including electric shock protection information, high-altitude protection information, environmental control lighting information, and wind and rain protection information.

[0018] Optionally, in the intelligent identification method for distribution workers' wearable devices described in an embodiment of the present application, the position detection of the worker image information is performed using the established multi-scale human body detection model, the safety factor information is supplemented according to the detected position information, and matched with the worker, and the worker and position and safety factor information are mapped and associated, including:

[0019] Process the images of personnel working on power distribution in the sample library and establish a multi-scale human detection model;

[0020] Performing position detection on the image information of the operator according to the multi-scale human body detection model to obtain position information of the operator, including operation position information and position positioning information;

[0021] Supplementing the electric shock protection information and the height protection information according to the operation position information;

[0022] Supplementing the environmental control lighting information and wind and rain protection information according to the position positioning information;

[0023] Matching the operator with the supplemented safety factor information;

[0024] The supplemented safety factor information is combined with the location information to map and associate the operator, and an association identifier of the operator, location, and safety factor is synthesized.

[0025] Optionally, in the intelligent identification method for distribution workers' wear described in an embodiment of the present application, the multiple detection results obtained based on multiple matching and association identification are integrated, and personnel attribute identification is performed to obtain the location area information and personnel attribute information to be tracked, and the standard wear requirement data of the workers is extracted from the distribution network operation wear database based on the location area information and personnel attribute information to be tracked, including:

[0026] Integrate multiple safety factor information obtained by multiple matching and association identifications with the location information of the corresponding operator to obtain the operator location safety factor information;

[0027] Identify the personnel attributes of the personnel in the personnel position safety factor information according to the collected position information combined with the safety factor information to obtain personnel attribute information;

[0028] synthesizing multiple location information of target workers corresponding to the personnel attribute information to obtain location area information to be tracked;

[0029] According to the personnel attribute information of the target operator and the information of the location area to be tracked, the preset distribution network operation wear database is input to extract the standard wear requirement data of the target operator.

[0030] Optionally, in the intelligent identification method for distribution workers' wearable devices described in an embodiment of the present application, the target worker's trajectory information is determined based on the intersection-and-union ratio of the target worker's tracking frame at each trajectory point. If the intersection-and-union ratio of the trajectory point is less than a preset threshold, it is determined that the target worker monitored at the trajectory point does not exist, and the trajectory point is removed, including:

[0031] Extracting each track point of the target operator according to the location area information to be tracked;

[0032] Obtaining an intersection-over-union ratio of the target worker tracking frame based on image processing of each track point;

[0033] Performing a threshold comparison based on the intersection-over-union ratio of each trajectory point and a preset threshold;

[0034] If the intersection-and-union ratio of the track point is less than a preset threshold, it is determined that the target worker does not exist in the track point and the track point is removed;

[0035] If the intersection-and-union ratio of the trajectory points is greater than a preset threshold, the target operator exists in the trajectory points, the trajectory points are valid, and the valid trajectory points are synthesized into the trajectory information of the target operator.

[0036] Optionally, in the intelligent identification method for distribution workers' wear according to the embodiment of the present application, recording the valid trajectory points in the target worker's trajectory information and the corresponding wear information, and determining the wear status of the target worker at each valid trajectory point according to the standard wear requirement data, includes:

[0037] Obtaining the wear information corresponding to the valid track points in the track information of the target operator, and recording it in combination with the corresponding track points;

[0038] Determine the wearing condition of the wear information of the trajectory point according to the standard wear requirement data corresponding to the target operator;

[0039] If the wearing information meets the standard wearing requirement data, the wearing condition of the track point is qualified.

[0040] Optionally, in the intelligent identification method for distribution workers' wearable devices described in an embodiment of the present application, the threshold judgment is performed based on the wearable device status of each worker in the preset monitoring time period, and the worker who exceeds the wearable device warning threshold is warned and recorded, including:

[0041] Obtain the wearing status of each operator at all valid track points within the preset monitoring period, and obtain the unqualified wearing data of the entire track;

[0042] Performing a threshold comparison between the full-track wearing failure data and a preset wearing warning threshold;

[0043] If the operator's full-track wearing unqualified data is greater than the wearing warning threshold, the operator will be warned and recorded.

[0044] In a second aspect, an embodiment of the present application provides an intelligent identification system for distribution workers' wearable clothing, the system comprising: a memory and a processor, the memory including a program for an intelligent identification method for distribution workers' wearable clothing, and the program for the intelligent identification method for distribution workers' wearable clothing, when executed by the processor, implementing the following steps:

[0045] Collect real-time monitoring videos of workers working on the distribution network without power outages, perform image recognition on the workers and their equipment in the monitoring videos, and obtain worker image information, equipment information, and safety factor information of the working environment;

[0046] The position of the operator image information is detected by the established multi-scale human body detection model, the safety factor information is supplemented according to the detected position information, and matched with the operator, and the operator, position and safety factor information are mapped and associated;

[0047] Based on multiple detection results obtained through multiple matching and association identification, information is integrated and person attributes are identified to obtain the location area information and person attribute information to be tracked;

[0048] Extracting standard wearing requirement data of the operator through the distribution network operation wear database according to the location area information to be tracked and the operator attribute information;

[0049] Determine the target worker's trajectory information based on the intersection-and-union ratio of the target worker's tracking frame at each trajectory point. If the intersection-and-union ratio of a trajectory point is less than a preset threshold, it is determined that the target worker monitored at the trajectory point does not exist, and the trajectory point is removed.

[0050] Recording the valid track points in the target worker's track information and the corresponding wearing information, and determining the wearing condition of the target worker at each valid track point according to the standard wearing requirement data;

[0051] A threshold judgment is made based on the full-trajectory wearing status of each operator within the preset monitoring time period, and operators who exceed the wearing warning threshold are warned and recorded.

[0052] Optionally, in the intelligent identification system for distribution workers' wearable devices described in the embodiments of the present application, the method of collecting real-time monitoring videos of distribution network workers without power outages, performing image recognition on the workers and their wearable devices in the monitoring videos, and obtaining worker image information, wearable device information, and safety factor information of the working environment includes:

[0053] Collect real-time monitoring video of the non-stop operation area of ​​the distribution network within the preset area;

[0054] Acquire images of workers in real-time surveillance videos, perform image recognition and image information extraction on workers and wearable devices, and obtain worker image information and wearable device information;

[0055] Based on image recognition, safety factor information of the power distribution operation environment is extracted, including electric shock protection information, high-altitude protection information, environmental control lighting information, and wind and rain protection information.

[0056] In a third aspect, an embodiment of the present application also provides a computer-readable storage medium, which includes a program for an intelligent identification method worn by distribution workers. When the program for the intelligent identification method worn by distribution workers is executed by a processor, the steps of the intelligent identification method worn by distribution workers as described in any one of the above items are implemented.

[0057] As can be seen from the above, the embodiments of the present application provide an intelligent identification method, system and medium for distribution workers' wear. By performing image recognition on workers and wearable equipment in monitoring videos, position detection is performed on worker image information, safety factor information is supplemented according to the position information, and matched and mapped with the worker for association identification, information integration is performed according to multiple detection results, and personnel attribute identification is performed to obtain the location area information and personnel attribute information to be tracked, and the standard wearing requirement data of the workers is extracted through the distribution network operation wear database, the trajectory information is judged according to the intersection and union ratio of each trajectory point, the wearing condition of each trajectory point is determined according to the standard wearing requirement data, and threshold judgment, warning and recording are performed according to the wearing condition of the entire trajectory; thereby, the distribution workers' wear is identified and positionally associated based on big data and distribution operation image recognition and analysis technology, and the wearing condition of the personnel trajectory is identified and evaluated according to the obtained standard wearing, so as to realize the identification and detection of the workers' wear, improve the effective identification and monitoring of the workers' wear, and ensure personnel safety.

[0058] Other features and advantages of the present application will be described in the following description and, in part, will become apparent from the description or be understood by practicing the embodiments of the present application. The objectives and other advantages of the present application can be achieved and obtained through the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0060] Figure 1 A flowchart of the intelligent identification method for distribution workers' wearable devices provided in an embodiment of the present application;

[0061] Figure 2 A flowchart of the intelligent identification method for distribution workers' clothing provided in an embodiment of the present application for obtaining worker image information, clothing information, and safety factor information of the working environment;

[0062] Figure 3 A flowchart of the intelligent identification method for distribution workers wearing clothing provided in an embodiment of the present application for supplementing and matching safety factor information and mapping associated identifiers;

[0063] Figure 4 A schematic diagram of the structure of the intelligent identification system worn by power distribution workers provided in an embodiment of the present application. DETAILED DESCRIPTION

[0064] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work fall within the scope of protection of the present application.

[0065] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0066] Please refer to Figure 1 , Figure 1 This is a flow chart of a method for intelligently identifying wearable devices for power distribution workers in some embodiments of the present application. The method is used in terminal devices such as computers and mobile phones. The method comprises the following steps:

[0067] S101. Collect real-time surveillance video of personnel working on the distribution network without power outages, perform image recognition on the personnel and their equipment in the surveillance video, and obtain information on the personnel's images, equipment, and safety factors of the working environment.

[0068] S102: Detect the position of the operator image information using the established multi-scale human body detection model, supplement the safety factor information based on the detected position information, match it with the operator, and map and associate the operator, position, and safety factor information;

[0069] S103: Integrate multiple detection results obtained through multiple matching and association identification, and perform person attribute recognition to obtain location area information and person attribute information to be tracked;

[0070] S104, extracting standard wearing requirement data of workers from the distribution network operation wear database according to the location area information to be tracked and the personnel attribute information;

[0071] S105: Determine the target worker's trajectory information based on the intersection-and-union ratio of the target worker's tracking frame at each trajectory point. If the intersection-and-union ratio of a trajectory point is less than a preset threshold, determine that the target worker monitored at the trajectory point does not exist, and remove the trajectory point.

[0072] S106, recording the valid track points in the target worker's track information and the corresponding wearing information, and determining the wearing condition of the target worker at each valid track point according to the standard wearing requirement data;

[0073] S107: Perform threshold judgment based on the full-track wearing conditions of each operator within a preset monitoring time period, and warn and record the operators who exceed the wearing warning threshold.

[0074] It should be noted that in order to realize the image recognition of the distribution operation environment, location and operators through video monitoring and image recognition, so as to identify the working environment and location as well as the corresponding personnel's work attributes, obtain the wearing standards that meet the distribution personnel's types and environment, and then track and identify the operators throughout the process through the wearing standards, determine the wearing conditions of the monitored operators throughout the process, and warn the operators who do not meet the wearing restrictions, so as to realize the safety supervision technology of the wearing conditions of the distribution operators based on image recognition and big data processing technology, and achieve intelligent safety management means to ensure the distribution construction. The specific process is: collect real-time monitoring videos of the distribution network's uninterrupted workers, perform image recognition on the operators and wearing equipment in the monitoring videos, obtain the operator's image information, wearing information and safety factor information of the working environment, and then perform multi-scale human body examination. The detection model performs position detection on the image information of the operator, supplements the safety factor information according to the detected position information, matches it with the operator, maps and associates the operator, position and safety factor information, integrates the information based on the multiple detection results, identifies the personnel attributes of the operator, obtains the location area information and personnel attribute information to be tracked, and then extracts the standard wearing requirement data of the operator through the distribution network operation wear database, judges the target operator's trajectory information according to the intersection and union ratio of the target operator tracking frame of each trajectory point, records the trajectory points and corresponding wearing information in the target operator's trajectory information, determines the wearing status of the target operator at each valid trajectory point according to the standard wearing requirement data, makes a threshold judgment based on the wearing status of each operator's entire trajectory, and warns and records the operator who exceeds the wearing warning threshold.

[0075] Please refer to Figure 2 , Figure 2 This is a flow chart of obtaining worker image information, wearable information, and safety factor information of the working environment through the wearable device worn by distribution workers in some embodiments of the present application. According to an embodiment of the present invention, the method of collecting real-time monitoring video of workers working on the distribution network without power outages, performing image recognition on the workers and wearable equipment in the monitoring video, and obtaining worker image information, wearable information, and safety factor information of the working environment is specifically as follows:

[0076] S201. Collect real-time monitoring video of the distribution network non-stop operation area within a preset region;

[0077] S202: Acquire an image of an operator in the real-time monitoring video, perform image recognition and image information extraction on the operator and the wearable device, and obtain operator image information and wearable device information;

[0078] S203. Recognize and extract safety factor information of the power distribution operation environment based on the image, including electric shock protection information, height protection information, environmental control lighting information, and wind and rain protection information.

[0079] It should be noted that in order to obtain the attire of personnel in the non-stop operation area within the monitored area, and to further determine the attire requirements based on the employee's ethnicity, work environment, and attire requirements, it is first necessary to obtain video images of the operation area, identify the images, and obtain information on personnel, attire, and environmental safety factors. That is, real-time monitoring video of the target area of ​​non-stop operation of the distribution network within a preset area is collected, and images containing personnel in the video are identified. The image information and attire information of the personnel are extracted from the images. The purpose is to identify the images to obtain image information containing personnel, and simultaneously identify the attire information of the personnel, such as the equipment and wearable devices such as helmets, protective clothing, insulating gloves, and protective boots worn. Then, image recognition is used to extract safety factor information of the distribution operation environment in which the personnel are located. Safety factor information is information on the safety protection elements of the distribution work, which is determined by identifying and determining the personnel information and work environment information in the images. It includes information on electric shock protection, high-altitude protection, environmental control lighting, and wind and rain protection. It reflects the protection and assurance elements required for personnel in the distribution environment, such as electric shock protection, high-altitude protection, lighting, and wind and rain protection.

[0080] Please refer to Figure 3 , Figure 3 This is a flow chart of the intelligent identification method for distribution workers in some embodiments of the present application, which supplements and matches safety factor information and maps associated identifiers. According to an embodiment of the present invention, the multi-scale human body detection model established is used to detect the position of the worker image information, supplements the safety factor information based on the detected position information, matches it with the worker, and maps associated identifiers between the worker, position, and safety factor information, specifically:

[0081] S301, processing the power distribution operation images of personnel in the sample library and establishing a multi-scale human body detection model;

[0082] S302: Performing position detection on the image information of the operator according to the multi-scale human body detection model to obtain position information of the operator, including operation position information and position positioning information;

[0083] S303, supplementing the electric shock protection information and the height protection information according to the operation position information;

[0084] S304: Supplement the environmental control lighting information and wind and rain protection information according to the position positioning information;

[0085] S305: Matching the operator with the supplemented safety factor information;

[0086] S306: Map and associate the supplemented safety factor information with the location information to the operator, and synthesize an association identifier of the operator, location, and safety factor.

[0087] It should be noted that in order to make the safety factor information required for the power distribution construction of the identified workers more accurate, the safety factor information needs to be supplemented according to the location information of the power distribution operation. Since the power distribution construction environment is often a large scene, and the workers are usually in a certain position of the power distribution unit, building, or facility, it is necessary to perform scene recognition based on the specific location of the workers, and supplement the safety factor information based on the construction location information of the scene recognition. The images of the power distribution operations of the workers in the sample library are processed and a multi-scale human body detection model is established. The sample library contains a large number of image samples of the power distribution operations of the workers. The multi-scale human body detection model is established through recognition training of a large number of image samples. The model can detect the information of the construction location of the workers, obtain the work location information and location positioning information, and then supplement the safety factor information respectively. According to the supplemented safety factor information, it is matched with the workers to make the construction safety factors and the workers accurately and effectively associated. Through a worker in a certain image, the safety factor information of his construction can be clearly identified, and then mapped and associated with the worker's location information to obtain the association between the worker, construction location and safety factors, and synthesize the associated identifier at the same time. Through matching and association, any supervised worker and his construction location are combined with the construction safety factors to achieve organic unity, which is convenient for further judgment of the protective wear of the supervised workers.

[0088] According to an embodiment of the present invention, the multiple detection results obtained based on multiple matching and association identification are integrated, and personnel attributes are identified to obtain the location area information and personnel attribute information to be tracked. Based on the location area information and personnel attribute information to be tracked, the standard wearing requirement data of the operating personnel is extracted through the distribution network operation wear database, specifically:

[0089] Integrate multiple safety factor information obtained by multiple matching and association identifications with the location information of the corresponding operator to obtain the operator location safety factor information;

[0090] Identify the personnel attributes of the personnel in the personnel position safety factor information according to the collected position information combined with the safety factor information to obtain personnel attribute information;

[0091] synthesizing multiple location information of target workers corresponding to the personnel attribute information to obtain location area information to be tracked;

[0092] According to the personnel attribute information of the target operator and the information of the location area to be tracked, the preset distribution network operation wear database is input to extract the standard wear requirement data of the target operator.

[0093] It should be noted that after identifying any worker, construction location and safety factor, when examining and judging the wearing situation of a certain worker, it is necessary to first clarify the nature of the worker's work or type of work, and obtain the worker's work area and construction trajectory points within the preset time, so as to determine whether the worker's wear meets the requirements of his job type and to supervise the worker's wear in the entire area. To achieve this design, the construction location and safety factor detection information of the worker obtained by multiple matching and associated identification are integrated to obtain a set of safety factor information of the worker at multiple construction locations. When a certain amount of information is collected, the construction locations and corresponding safety factor information where the worker often appears can be identified through the information collection, and the worker attributes can be clarified to obtain the personnel attribute information. For example, through the collected information, it is found that a worker often appears in front of the detection box of the distribution network, wearing the wearable equipment and tools of the detection personnel, so it can be identified that the worker's type of work belongs to The line inspection personnel then synthesizes the multiple construction locations of the identified target operator into the location area information to be tracked based on the personnel attribute information, and obtains the construction location area of ​​the target operator, such as the area where a line inspection personnel has stayed during construction within a week. By clarifying the personnel attribute information of the target operator and the location area information to be tracked, that is, obtaining the type of work attribute and construction area of ​​the operator, the standard wearing requirement data of the target operator can be obtained by combining the construction location and area with the personnel attributes in the preset distribution network operation wear database, that is, querying the database to find out the standard wearing requirements that the target operator should wear in his type of work, construction location and area. For example, if a line inspection personnel performs line inspection work on the top of a telephone pole in a rural field in the north on a winter night, then his standard wearing requirements need to include standard wearing equipment, devices, tools, etc. for cold protection, lighting, warning, positioning, anti-fall, and anti-electric shock, and the standard wearing requirement data is set according to his standard wearing requirements.

[0094] According to an embodiment of the present invention, the target worker's trajectory information is determined based on the intersection-and-union ratio of the target worker's tracking frame at each trajectory point. If the intersection-and-union ratio of a trajectory point is less than a preset threshold, it is determined that the target worker monitored at the trajectory point does not exist, and the trajectory point is removed. Specifically,

[0095] Extracting each track point of the target operator according to the location area information to be tracked;

[0096] Obtaining an intersection-over-union ratio of the target worker tracking frame based on image processing of each track point;

[0097] Performing a threshold comparison based on the intersection-over-union ratio of each trajectory point and a preset threshold;

[0098] If the intersection-and-union ratio of the track point is less than a preset threshold, it is determined that the target worker does not exist in the track point and the track point is removed;

[0099] If the intersection-and-union ratio of the trajectory points is greater than a preset threshold, the target operator exists in the trajectory points, the trajectory points are valid, and the valid trajectory points are synthesized into the trajectory information of the target operator.

[0100] It should be noted that due to the problem of capture and recognition deviation in the trajectory of the target workers in the construction area within the preset time period, the target workers do not appear in the trajectories of some areas. Therefore, in order to accurately grasp the accurate construction trajectory points of the target workers, it is necessary to perform image processing on each trajectory point to judge the accuracy of capturing the target workers. The frame intersection and union ratio obtained by image processing of each trajectory point of the extracted target workers is used to perform a preset threshold judgment. If the intersection and union ratio of a trajectory point is less than the preset threshold, it is determined that the target worker at the trajectory point does not exist, and the trajectory point is an invalid trajectory point and is removed. On the contrary, if it is greater than the preset threshold, the trajectory point is valid and the valid trajectory point is judged. Synthesis, obtain the trajectory information of the target operator. The intersection over union (IoU) of the target operator's tracking frame is a standard for measuring the accuracy of object / person detection. According to the real area of ​​a target marked in the image, the predicted area of ​​the target is obtained in the target detection. The detection accuracy is evaluated by calculating the value of the intersection over union. Generally, the larger the intersection over union value, the more accurate the measurement result. The intersection over union is calculated as follows: IoU = intersection of detection area and real area / union of detection area and real area. In this scheme, the intersection over union of the tracking frame obtained by image processing determines whether the target operator is actually present in the detected area, so as to improve the accuracy of target operator detection and filter out invalid trajectory points in the trajectory information of the tracking area.

[0101] According to an embodiment of the present invention, the valid track points in the target worker's track information and the corresponding wearing information are recorded, and the wearing condition of the target worker at each valid track point is determined according to the standard wearing requirement data, specifically:

[0102] Obtaining the wear information corresponding to the valid track points in the track information of the target operator, and recording it in combination with the corresponding track points;

[0103] Determine the wearing condition of the wear information of the trajectory point according to the standard wear requirement data corresponding to the target operator;

[0104] If the wearing information meets the standard wearing requirement data, the wearing condition of the track point is qualified.

[0105] It should be noted that the wearing information corresponding to the valid track points in the obtained trajectory information of the target operator is compared with the corresponding standard wearing requirement data to determine the wearing condition of the target operator at the track point. If the wearing information of a certain track point meets the preset comparison requirements of the standard wearing requirement data, it is determined that the wearing of the target operator at the track point meets the standard wearing requirement data, and the wearing condition of the track point is qualified. The wearing condition of each track point in the trajectory of the target operator is determined in turn.

[0106] According to an embodiment of the present invention, the threshold judgment is performed based on the full-track wearing status of each operator within a preset monitoring time period, and the operator who exceeds the wearing warning threshold is warned and recorded, specifically:

[0107] Obtain the wearing status of each operator at all valid track points within the preset monitoring period, and obtain the unqualified wearing data of the entire track;

[0108] Performing a threshold comparison between the full-track wearing failure data and a preset wearing warning threshold;

[0109] If the operator's full-track wearing unqualified data is greater than the wearing warning threshold, the operator will be warned and recorded.

[0110] It should be noted that after identifying the wearing conditions of all valid trajectory points of each monitored operator within the preset monitoring time period, the unqualified wearing conditions of each operator in the entire trajectory, that is, the unqualified wearing data, can be obtained. The unqualified wearing data of the entire trajectory can be compared with the preset wearing warning threshold. If the unqualified wearing data of the operator in the entire trajectory is greater than the wearing warning threshold, the operator is warned and recorded. That is, by comparing the unqualified wearing data of a certain operator in the entire trajectory with the preset wearing threshold, the wearing condition of the operator throughout the operation is judged. If the unqualified wearing data exceeds the preset threshold requirement, it indicates that the operator has worn non-compliant clothing more times than the standard during the entire operation within the time period, and the operator needs to be warned and recorded. By judging the wearing condition of the entire trajectory, the person who violates the wearing regulations can be identified, warned and dealt with, so as to avoid safety operation accidents caused by non-compliant wearing, so as to realize the safety supervision technology of the wearing condition of distribution operators based on image recognition and big data processing technology in this solution, and achieve intelligent safety management means to ensure distribution construction.

[0111] like Figure 4As shown, the present invention also discloses an intelligent identification system for distribution workers, including a memory 41 and a processor 42. The memory includes an intelligent identification method program for distribution workers, and when the processor executes the intelligent identification method program for distribution workers, the following steps are implemented when the abnormal vital sign correction data is executed:

[0112] Collect real-time monitoring videos of workers working on the distribution network without power outages, perform image recognition on the workers and their equipment in the monitoring videos, and obtain worker image information, equipment information, and safety factor information of the working environment;

[0113] The position of the operator image information is detected by the established multi-scale human body detection model, the safety factor information is supplemented according to the detected position information, and matched with the operator, and the operator, position and safety factor information are mapped and associated;

[0114] Based on multiple detection results obtained through multiple matching and association identification, information is integrated and person attributes are identified to obtain the location area information and person attribute information to be tracked;

[0115] Extracting standard wearing requirement data of the operator through the distribution network operation wear database according to the location area information to be tracked and the operator attribute information;

[0116] Determine the target worker's trajectory information based on the intersection-and-union ratio of the target worker's tracking frame at each trajectory point. If the intersection-and-union ratio of a trajectory point is less than a preset threshold, it is determined that the target worker monitored at the trajectory point does not exist, and the trajectory point is removed.

[0117] Recording the valid track points in the target worker's track information and the corresponding wearing information, and determining the wearing condition of the target worker at each valid track point according to the standard wearing requirement data;

[0118] A threshold judgment is made based on the full-trajectory wearing status of each operator within the preset monitoring time period, and operators who exceed the wearing warning threshold are warned and recorded.

[0119] It should be noted that in order to realize the image recognition of the distribution operation environment, location and operators through video monitoring and image recognition, so as to identify the working environment and location as well as the corresponding personnel's work attributes, obtain the wearing standards that meet the distribution personnel's types and environment, and then track and identify the operators throughout the process through the wearing standards, determine the wearing conditions of the monitored operators throughout the process, and warn the operators who do not meet the wearing restrictions, so as to realize the safety supervision technology of the wearing conditions of the distribution operators based on image recognition and big data processing technology, and achieve intelligent safety management means to ensure the distribution construction. The specific process is: collect real-time monitoring videos of the distribution network's uninterrupted workers, perform image recognition on the operators and wearing equipment in the monitoring videos, obtain the operator's image information, wearing information and safety factor information of the working environment, and then perform multi-scale human body examination. The detection model performs position detection on the image information of the operator, supplements the safety factor information according to the detected position information, matches it with the operator, maps and associates the operator, position and safety factor information, integrates the information based on the multiple detection results, identifies the personnel attributes of the operator, obtains the location area information and personnel attribute information to be tracked, and then extracts the standard wearing requirement data of the operator through the distribution network operation wear database, judges the target operator's trajectory information according to the intersection and union ratio of the target operator tracking frame of each trajectory point, records the trajectory points and corresponding wearing information in the target operator's trajectory information, determines the wearing status of the target operator at each valid trajectory point according to the standard wearing requirement data, makes a threshold judgment based on the wearing status of each operator's entire trajectory, and warns and records the operator who exceeds the wearing warning threshold.

[0120] According to an embodiment of the present invention, the real-time monitoring video of the distribution network uninterrupted operation personnel is collected, and image recognition of the operating personnel and the equipment they wear in the monitoring video is performed to obtain the image information of the operating personnel, the information of the equipment they wear, and the safety factor information of the working environment, specifically:

[0121] Collect real-time monitoring video of the non-stop operation area of ​​the distribution network within the preset area;

[0122] Acquire images of workers in real-time surveillance videos, perform image recognition and image information extraction on workers and wearable devices, and obtain worker image information and wearable device information;

[0123] Based on image recognition, safety factor information of the power distribution operation environment is extracted, including electric shock protection information, high-altitude protection information, environmental control lighting information, and wind and rain protection information.

[0124] It should be noted that in order to obtain the attire of personnel in the non-stop operation area within the monitored area, and to further determine the attire requirements based on the employee's ethnicity, work environment, and attire requirements, it is first necessary to obtain video images of the operation area, identify the images, and obtain information on personnel, attire, and environmental safety factors. That is, real-time monitoring video of the target area of ​​non-stop operation of the distribution network within a preset area is collected, and images containing personnel in the video are identified. The image information and attire information of the personnel are extracted from the images. The purpose is to identify the images to obtain image information containing personnel, and simultaneously identify the attire information of the personnel, such as the equipment and wearable devices such as helmets, protective clothing, insulating gloves, and protective boots worn. Then, image recognition is used to extract safety factor information of the distribution operation environment in which the personnel are located. Safety factor information is information on the safety protection elements of the distribution work, which is determined by identifying and determining the personnel information and work environment information in the images. It includes information on electric shock protection, high-altitude protection, environmental control lighting, and wind and rain protection. It reflects the protection and assurance elements required for personnel in the distribution environment, such as electric shock protection, high-altitude protection, lighting, and wind and rain protection.

[0125] According to an embodiment of the present invention, the position of the operator image information is detected by the established multi-scale human body detection model, the safety factor information is supplemented according to the detected position information, and matched with the operator, and the operator and position and safety factor information are mapped and associated. Specifically,

[0126] Process the images of personnel working on power distribution in the sample library and establish a multi-scale human detection model;

[0127] Performing position detection on the image information of the operator according to the multi-scale human body detection model to obtain position information of the operator, including operation position information and position positioning information;

[0128] Supplementing the electric shock protection information and the height protection information according to the operation position information;

[0129] Supplementing the environmental control lighting information and wind and rain protection information according to the position positioning information;

[0130] Matching the operator with the supplemented safety factor information;

[0131] The supplemented safety factor information is combined with the location information to map and associate the operator, and an association identifier of the operator, location, and safety factor is synthesized.

[0132] It should be noted that in order to make the safety factor information required for the power distribution construction of the identified workers more accurate, the safety factor information needs to be supplemented according to the location information of the power distribution operation. Since the power distribution construction environment is often a large scene, and the workers are usually in a certain position of the power distribution unit, building, or facility, it is necessary to perform scene recognition based on the specific location of the workers, and supplement the safety factor information based on the construction location information of the scene recognition. The images of the power distribution operations of the workers in the sample library are processed and a multi-scale human body detection model is established. The sample library contains a large number of image samples of the power distribution operations of the workers. The multi-scale human body detection model is established through recognition training of a large number of image samples. The model can detect the information of the construction location of the workers, obtain the work location information and location positioning information, and then supplement the safety factor information respectively. According to the supplemented safety factor information, it is matched with the workers to make the construction safety factors and the workers accurately and effectively associated. Through a worker in a certain image, the safety factor information of his construction can be clearly identified, and then mapped and associated with the worker's location information to obtain the association between the worker, construction location and safety factors, and synthesize the associated identifier at the same time. Through matching and association, any supervised worker and his construction location are combined with the construction safety factors to achieve organic unity, which is convenient for further judgment of the protective wear of the supervised workers.

[0133] According to an embodiment of the present invention, the multiple detection results obtained based on multiple matching and association identification are integrated, and personnel attributes are identified to obtain the location area information and personnel attribute information to be tracked. Based on the location area information and personnel attribute information to be tracked, the standard wearing requirement data of the operating personnel is extracted through the distribution network operation wear database, specifically:

[0134] Integrate multiple safety factor information obtained by multiple matching and association identifications with the location information of the corresponding operator to obtain the operator location safety factor information;

[0135] Identify the personnel attributes of the personnel in the personnel position safety factor information according to the collected position information combined with the safety factor information to obtain personnel attribute information;

[0136] synthesizing multiple location information of target workers corresponding to the personnel attribute information to obtain location area information to be tracked;

[0137] According to the personnel attribute information of the target operator and the information of the location area to be tracked, the preset distribution network operation wear database is input to extract the standard wear requirement data of the target operator.

[0138] It should be noted that after identifying any worker, construction location and safety factor, when examining and judging the wearing situation of a certain worker, it is necessary to first clarify the nature of the worker's work or type of work, and obtain the worker's work area and construction trajectory points within the preset time, so as to determine whether the worker's wear meets the requirements of his job type and to supervise the worker's wear in the entire area. To achieve this design, the construction location and safety factor detection information of the worker obtained by multiple matching and associated identification are integrated to obtain a set of safety factor information of the worker at multiple construction locations. When a certain amount of information is collected, the construction locations and corresponding safety factor information where the worker often appears can be identified through the information collection, and the worker attributes can be clarified to obtain the personnel attribute information. For example, through the collected information, it is found that a worker often appears in front of the detection box of the distribution network, wearing the wearable equipment and tools of the detection personnel, so it can be identified that the worker's type of work belongs to The line inspection personnel then synthesizes the multiple construction locations of the identified target operator into the location area information to be tracked based on the personnel attribute information, and obtains the construction location area of ​​the target operator, such as the area where a line inspection personnel has stayed during construction within a week. By clarifying the personnel attribute information of the target operator and the location area information to be tracked, that is, obtaining the type of work attribute and construction area of ​​the operator, the standard wearing requirement data of the target operator can be obtained by combining the construction location and area with the personnel attributes in the preset distribution network operation wear database, that is, querying the database to find out the standard wearing requirements that the target operator should wear in his type of work, construction location and area. For example, if a line inspection personnel performs line inspection work on the top of a telephone pole in a rural field in the north on a winter night, then his standard wearing requirements need to include standard wearing equipment, devices, tools, etc. for cold protection, lighting, warning, positioning, anti-fall, and anti-electric shock, and the standard wearing requirement data is set according to his standard wearing requirements.

[0139] According to an embodiment of the present invention, the target worker's trajectory information is determined based on the intersection-and-union ratio of the target worker's tracking frame at each trajectory point. If the intersection-and-union ratio of a trajectory point is less than a preset threshold, it is determined that the target worker monitored at the trajectory point does not exist, and the trajectory point is removed. Specifically,

[0140] Extracting each track point of the target operator according to the location area information to be tracked;

[0141] Obtaining an intersection-over-union ratio of the target worker tracking frame based on image processing of each track point;

[0142] Performing a threshold comparison based on the intersection-over-union ratio of each trajectory point and a preset threshold;

[0143] If the intersection-and-union ratio of the track point is less than a preset threshold, it is determined that the target worker does not exist in the track point and the track point is removed;

[0144] If the intersection-and-union ratio of the trajectory points is greater than a preset threshold, the target operator exists in the trajectory points, the trajectory points are valid, and the valid trajectory points are synthesized into the trajectory information of the target operator.

[0145] It should be noted that due to the problem of capture and recognition deviation in the trajectory of the target workers in the construction area within the preset time period, the target workers do not appear in the trajectories of some areas. Therefore, in order to accurately grasp the accurate construction trajectory points of the target workers, it is necessary to perform image processing on each trajectory point to judge the accuracy of capturing the target workers. The frame intersection and union ratio obtained by image processing of each trajectory point of the extracted target workers is used to perform a preset threshold judgment. If the intersection and union ratio of a trajectory point is less than the preset threshold, it is determined that the target worker at the trajectory point does not exist, and the trajectory point is an invalid trajectory point and is removed. On the contrary, if it is greater than the preset threshold, the trajectory point is valid and the valid trajectory point is judged. Synthesis, obtain the trajectory information of the target operator. The intersection over union (IoU) of the target operator's tracking frame is a standard for measuring the accuracy of object / person detection. According to the real area of ​​a target marked in the image, the predicted area of ​​the target is obtained in the target detection. The detection accuracy is evaluated by calculating the value of the intersection over union. Generally, the larger the intersection over union value, the more accurate the measurement result. The intersection over union is calculated as follows: IoU = intersection of detection area and real area / union of detection area and real area. In this scheme, the intersection over union of the tracking frame obtained by image processing determines whether the target operator is actually present in the detected area, so as to improve the accuracy of target operator detection and filter out invalid trajectory points in the trajectory information of the tracking area.

[0146] According to an embodiment of the present invention, the valid track points in the target worker's track information and the corresponding wearing information are recorded, and the wearing condition of the target worker at each valid track point is determined according to the standard wearing requirement data, specifically:

[0147] Obtaining the wear information corresponding to the valid track points in the track information of the target operator, and recording it in combination with the corresponding track points;

[0148] Determine the wearing condition of the wear information of the trajectory point according to the standard wear requirement data corresponding to the target operator;

[0149] If the wearing information meets the standard wearing requirement data, the wearing condition of the track point is qualified.

[0150] It should be noted that the wearing information corresponding to the valid track points in the obtained trajectory information of the target operator is compared with the corresponding standard wearing requirement data to determine the wearing condition of the target operator at the track point. If the wearing information of a certain track point meets the preset comparison requirements of the standard wearing requirement data, it is determined that the wearing of the target operator at the track point meets the standard wearing requirement data, and the wearing condition of the track point is qualified. The wearing condition of each track point in the trajectory of the target operator is determined in turn.

[0151] According to an embodiment of the present invention, the threshold judgment is performed based on the full-track wearing status of each operator within a preset monitoring time period, and the operator who exceeds the wearing warning threshold is warned and recorded, specifically:

[0152] Obtain the wearing status of each operator at all valid track points within the preset monitoring period, and obtain the unqualified wearing data of the entire track;

[0153] Performing a threshold comparison between the full-track wearing failure data and a preset wearing warning threshold;

[0154] If the operator's full-track wearing unqualified data is greater than the wearing warning threshold, the operator will be warned and recorded.

[0155] It should be noted that after identifying the wearing conditions of all valid trajectory points of each monitored operator within the preset monitoring time period, the unqualified wearing conditions of each operator in the entire trajectory, that is, the unqualified wearing data, can be obtained. The unqualified wearing data of the entire trajectory can be compared with the preset wearing warning threshold. If the unqualified wearing data of the operator in the entire trajectory is greater than the wearing warning threshold, the operator is warned and recorded. That is, by comparing the unqualified wearing data of a certain operator in the entire trajectory with the preset wearing threshold, the wearing condition of the operator throughout the operation is judged. If the unqualified wearing data exceeds the preset threshold requirement, it indicates that the operator has worn non-compliant clothing more times than the standard during the entire operation within the time period, and the operator needs to be warned and recorded. By judging the wearing condition of the entire trajectory, the person who violates the wearing regulations can be identified, warned and dealt with, so as to avoid safety operation accidents caused by non-compliant wearing, so as to realize the safety supervision technology of the wearing condition of distribution operators based on image recognition and big data processing technology in this solution, and achieve intelligent safety management means to ensure distribution construction.

[0156] The third aspect of the present invention provides a readable storage medium, which includes a program for an intelligent identification method worn by distribution workers. When the program for the intelligent identification method worn by distribution workers is executed by a processor, the steps of the intelligent identification method worn by distribution workers as described in any one of the above items are implemented.

[0157] The present invention discloses an intelligent identification method, system and medium for distribution workers' wear. The method performs image recognition on workers and wearable equipment in monitoring videos, performs position detection on workers' image information, supplements safety factor information based on the position information, matches and maps associated identifiers with workers, integrates information based on multiple detection results, and performs personnel attribute recognition to obtain location area information and personnel attribute information to be tracked. The method also extracts workers' standard wear requirement data from a distribution network operation wear database, judges trajectory information based on the intersection-union ratio of each trajectory point, determines the wear condition of each trajectory point based on the standard wear requirement data, and performs threshold judgment, warning and recording based on the wear condition of the entire trajectory. The method thus identifies and associates the wear of distribution workers with their positions based on big data and distribution operation image recognition and analysis technology, and identifies and evaluates the wear condition of the personnel trajectory based on the obtained standard wear, thereby realizing identification and detection of workers' wear, improving effective identification and monitoring of workers' wear, and ensuring personnel safety.

[0158] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0159] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed across multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0160] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0161] Those skilled in the art will understand that all or part of the steps of the above-mentioned method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiment; and the aforementioned storage medium includes: mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, etc. Various media that can store program codes.

[0162] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

Claims

1. An intelligent identification method for distribution workers' wear, characterized in that: The following steps are involved: Collect real-time monitoring videos of workers working on the distribution network without power outages, perform image recognition on the workers and their equipment in the monitoring videos, and obtain worker image information, equipment information, and safety factor information of the working environment; The position of the operator image information is detected by the established multi-scale human body detection model, the safety factor information is supplemented according to the detected position information, and matched with the operator, and the operator, position and safety factor information are mapped and associated; Based on multiple detection results obtained through multiple matches and association identifications, information is integrated and person attributes are identified to obtain the location area information and person attribute information to be tracked, including: Integrate multiple safety factor information obtained by multiple matching and association identifications with the location information of the corresponding operator to obtain the operator location safety factor information; Identify the personnel attributes of the personnel in the personnel position safety factor information according to the collected position information combined with the safety factor information to obtain personnel attribute information; synthesizing multiple location information of target workers corresponding to the personnel attribute information to obtain location area information to be tracked; Extracting standard wearing requirement data of the operator through the distribution network operation wear database according to the location area information to be tracked and the operator attribute information; The target worker's trajectory information is determined based on the intersection over union (IoU) of the target worker's tracking frame at each trajectory point. If the IoU of a trajectory point is less than a preset threshold, it is determined that the target worker monitored at the trajectory point does not exist and the trajectory point is removed. The IoU is calculated as follows: IoU = Intersection of Detected Area and Ground Truth Area / Union of Detected Area and Ground Truth Area. Recording the valid track points in the target worker's track information and the corresponding wearing information, and determining the wearing condition of the target worker at each valid track point according to the standard wearing requirement data; A threshold judgment is made based on the full-trajectory wearing status of each operator within the preset monitoring time period, and operators who exceed the wearing warning threshold are warned and recorded.

2. The intelligent identification method for distribution workers according to claim 1 is characterized in that: The method collects real-time monitoring videos of distribution network operators without power outages, performs image recognition on the operators and their equipment in the monitoring videos, and obtains operator image information, equipment information, and safety factor information of the working environment, including: Collect real-time monitoring video of the non-stop operation area of ​​the distribution network within the preset area; Acquire images of workers in real-time surveillance videos, perform image recognition and image information extraction on workers and wearable devices, and obtain worker image information and wearable device information; Based on image recognition, safety factor information of the power distribution operation environment is extracted, including electric shock protection information, high-altitude protection information, environmental control lighting information, and wind and rain protection information.

3. The intelligent identification method for distribution workers according to claim 2 is characterized in that: The multi-scale human body detection model is used to detect the position of the operator image information, and the safety factor information is supplemented according to the detected position information, and matched with the operator, and the operator and the position and safety factor information are mapped and associated, including: Process the images of personnel working on power distribution in the sample library and establish a multi-scale human body detection model; Performing position detection on the operator image information according to the multi-scale human body detection model to obtain the operator's position information, including operation position information and position positioning information; Supplementing the electric shock protection information and the height protection information according to the operation position information; Supplementing the environmental control lighting information and wind and rain protection information according to the position positioning information; Matching the operator with the supplemented safety factor information; The supplemented safety factor information is combined with the location information to map and associate the operator, and an association identifier of the operator, location, and safety factor is synthesized.

4. The intelligent identification method for distribution workers according to claim 1, characterized in that: According to the location area information to be tracked and the personnel attribute information, standard wearing requirement data of the operating personnel is extracted through the distribution network operation wearing database, including: According to the personnel attribute information of the target operator and the information of the location area to be tracked, the preset distribution network operation wear database is input to extract the standard wear requirement data of the target operator.

5. The intelligent identification method for distribution workers' wear according to claim 4 is characterized in that: The target worker trajectory information is determined based on the intersection-and-union ratio of the target worker tracking frame of each trajectory point. If the intersection-and-union ratio of the trajectory point is less than a preset threshold, it is determined that the target worker monitored at the trajectory point does not exist, and the trajectory point is removed, including: Extracting each track point of the target operator according to the location area information to be tracked; Obtaining an intersection-over-union ratio of the target worker tracking frame based on image processing of each track point; Performing a threshold comparison based on the intersection-over-union ratio of each trajectory point and a preset threshold; If the intersection-and-union ratio of the track point is less than a preset threshold, it is determined that the target worker does not exist in the track point and the track point is removed; If the intersection-and-union ratio of the trajectory points is greater than a preset threshold, the target operator exists in the trajectory points, the trajectory points are valid, and the valid trajectory points are synthesized into the trajectory information of the target operator.

6. The intelligent identification method for distribution workers' wear according to claim 5, characterized in that: The recording of the valid track points in the target worker's track information and the corresponding wearing information, and determining the wearing condition of the target worker at each valid track point according to the standard wearing requirement data, includes: Obtaining the wear information corresponding to the valid track points in the track information of the target operator, and recording it in combination with the corresponding track points; Determine the wearing condition of the wear information of the trajectory point according to the standard wear requirement data corresponding to the target operator; If the wearing information meets the standard wearing requirement data, the wearing condition of the track point is qualified.

7. The intelligent identification method for distribution workers' wear according to claim 6, characterized in that: The threshold judgment is performed based on the full-track wearing status of each operator within the preset monitoring time period, and the operator who exceeds the wearing warning threshold is warned and recorded, including: Obtain the wearing status of each operator at all valid track points within the preset monitoring period, and obtain the unqualified wearing data of the entire track; Performing a threshold comparison between the full-track wearing failure data and a preset wearing warning threshold; If the operator's full-track wearing unqualified data is greater than the wearing warning threshold, the operator will be warned and recorded.

8. An intelligent identification system worn by power distribution workers, characterized in that: The system includes: a memory and a processor, wherein the memory includes a program for an intelligent identification method for distribution workers wearing clothing, and when the program is executed by the processor, the following steps are implemented: Collect real-time monitoring videos of workers working on the distribution network without power outages, perform image recognition on the workers and their equipment in the monitoring videos, and obtain worker image information, equipment information, and safety factor information of the working environment; The position of the operator image information is detected by the established multi-scale human body detection model, the safety factor information is supplemented according to the detected position information, and matched with the operator, and the operator, position and safety factor information are mapped and associated; Based on multiple detection results obtained through multiple matches and association identifications, information is integrated and person attributes are identified to obtain the location area information and person attribute information to be tracked, including: Integrate multiple safety factor information obtained by multiple matching and association identifications with the location information of the corresponding operator to obtain the operator location safety factor information; Identify the personnel attributes of the personnel in the personnel position safety factor information according to the collected position information combined with the safety factor information to obtain personnel attribute information; synthesizing multiple location information of target workers corresponding to the personnel attribute information to obtain location area information to be tracked; Extracting standard wearing requirement data of the operator through the distribution network operation wear database according to the location area information to be tracked and the operator attribute information; The target worker's trajectory information is determined based on the intersection over union (IoU) of the target worker's tracking frame at each trajectory point. If the IoU of a trajectory point is less than a preset threshold, it is determined that the target worker monitored at the trajectory point does not exist and the trajectory point is removed. The IoU is calculated as follows: IoU = Intersection of Detected Area and Ground Truth Area / Union of Detected Area and Ground Truth Area. Recording the valid track points in the target worker's track information and the corresponding wearing information, and determining the wearing condition of the target worker at each valid track point according to the standard wearing requirement data; A threshold judgment is made based on the full-trajectory wearing status of each operator within the preset monitoring time period, and operators who exceed the wearing warning threshold are warned and recorded.

9. The intelligent identification system for distribution workers according to claim 8, characterized in that: The method collects real-time monitoring videos of distribution network operators without power outages, performs image recognition on the operators and their equipment in the monitoring videos, and obtains operator image information, equipment information, and safety factor information of the working environment, including: Collect real-time monitoring video of the non-stop operation area of ​​the distribution network within the preset area; Acquire images of workers in real-time surveillance videos, perform image recognition and image information extraction on workers and wearable devices, and obtain worker image information and wearable device information; Based on image recognition, safety factor information of the power distribution operation environment is extracted, including electric shock protection information, high-altitude protection information, environmental control lighting information, and wind and rain protection information.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium includes an intelligent identification method program for distribution workers' wear. When the intelligent identification method program for distribution workers' wear is executed by a processor, the steps of the intelligent identification method for distribution workers' wear as described in any one of claims 1 to 7 are implemented.

Images